JP4408355B2 - Image processing apparatus and image processing program - Google Patents

Description

The present invention relates to an image processing apparatus, it relates to an image processing method and image processing program, and more particularly, to an image processing instrumentation 置及 beauty image processing program for performing image processing for grasping the input and the status of the human for a given area.

  It is very important to evaluate the operational status of various facilities such as hospitals, libraries, government offices, and theaters in order for the facility to function well. It is necessary to grasp the movement situation.

  In this method, facility users receive non-contact wireless transmitters such as RFID (Radio Frequency Identification) tags and receive them, and receive the radio waves from the transmitters at the necessary locations in the facility. A method is conceivable in which a machine is installed in advance and the movement status of the user is grasped based on the reception result of each receiver.

  However, this method is not practical because it takes time and effort to distribute and collect transmitters to each facility user, and it also takes time and cost to install and collect receivers.

  Therefore, as a technique that can be applied to grasp the movement status of facility users simply and at low cost, Patent Document 1 is provided with a camera for photographing a passage through which a person passes, and a plurality of cameras along the passage. The number of people crossing the area within the unit time is counted for each monitoring area in the image data obtained by the above camera shooting, and the variation between the areas is within a certain range. If there is a monitoring area where the average value of the total number of people in all monitoring areas is the true total number of people and the variation exceeds a certain range, it is determined that a disturbance has occurred in the screen of that area. A technique for calculating the true total number of people based on the total number of people in the monitoring area excluding the above is disclosed.

According to this technology, it is possible to grasp the number of people passing through a given passage in the facility, so it is possible to grasp the overall movement status of facility users by implementing this technology for multiple passages in the facility. can do.
JP 2002-74371 A

  By the way, as a part of grasping the movement situation in the facility by the facility user, there is a demand for grasping the user entry / exit situation with respect to a predetermined room in the facility. For example, when the facility is a hospital, the examination rooms such as internal medicine and surgery can be exemplified as the predetermined room. In this case, grasping the number of patients entering and leaving the examination room and knowing the stay period (examination period) of each patient in the examination room is very important in evaluating the operating status of the examination room. Is important to.

  However, with the technique of Patent Document 1, although the number of people entering and leaving the room can be grasped, the stay period cannot be grasped, and the movement status of facility users cannot be grasped sufficiently. There was a problem that.

  That is, in the technique of Patent Document 1, the number of people crossing the monitoring area is counted based on the change in the image, and the person who crosses the area cannot be specified. I can't do it.

The present invention has been made to solve the above problems, and an object thereof is to provide an image processing instrumentation 置及 beauty image processing program that can recognize the moving condition of the facility users at a high level.

In order to achieve the above object, an image processing apparatus according to claim 1 is an image processing apparatus for performing image processing for grasping a person's entry / exit status with respect to a predetermined area, and entering the predetermined area and the predetermined area. An image acquisition means for capturing a plurality of pieces of image information by photographing a passage area through which a person passes when leaving the area at predetermined intervals; a plurality of pieces of image information acquired by the image acquisition means; and the passage area Compare with background image information predetermined as image information acquired by the image acquisition means in a state of not passing through, and from the plurality of image information, a time series of difference areas on the image with the background image information A person passes through the passing area to enter the predetermined area with the image information substantially matching the passing direction when the person passes through the passing area so that the moving direction in order enters the predetermined area. When Is extracted as the first image information obtained by photographing, leaving the image information substantially coincides with the passage direction when the moving direction is the passage of the passage region person to exit from the predetermined area from the predetermined area Therefore, the first extraction means for extracting as the second image information acquired by photographing when a person is passing through the passage area, and the newest first image information extracted by the first extraction means In the direction toward the old first image information, the first image extracted by the first extraction unit in the order of the oldest second image information from the newest second image information extracted by the first extraction unit. to extract the second image information and the first image information representing an image including irradiation combined and the same person and the extracted second image information by one image information and the first extracting means And it includes a second extracting means, a storage means for storing the acquisition time by each of the image acquisition unit of the first image information extracted and the second image data by the second extraction means.

  According to the image processing apparatus of claim 1, the image acquisition unit acquires a plurality of pieces of image information by photographing a passing area through which a person passes when entering a predetermined area and exiting the predetermined area every predetermined period. Is done.

  The passing area may be a single area, or the area through which a person passes when entering a predetermined area is different from the area through which a person passes when exiting the predetermined area, or at least of these areas There may be a plurality of regions, such as a case where a plurality of ones exist. If the passage area is a single area, only one image acquisition unit needs to be provided. Even in the case of a plurality of areas, only one image acquisition unit is provided if each area is located relatively close. Just do it. In these cases, the cost can be reduced as compared with the case where a plurality of image acquisition means are provided, and the processing can be simplified. The image acquisition means includes a digital camera such as a digital electronic still camera or a digital video camera that acquires digital image information of a subject image by photographing.

In the present invention, the first extraction unit Thus, as image information a person the passage region and a plurality of image information acquired is acquired by the image acquisition means in a state that has not passed by the image acquisition unit Compared with the predetermined background image information, and from the plurality of image information , the moving area in the time series order of the difference area on the image with the background image information enters the predetermined area, The image information substantially coincident with the passing direction when the person passes is extracted as first image information acquired by photographing when the person is passing through the passing area in order to enter the predetermined area , the image information passing substantially coincides with the direction the person the passage region to exit from the predetermined area is passing through when the moving direction passes person the passage region to exit from the predetermined area It is extracted as the second image information obtained by Kino shooting, to thus in a direction toward the oldest first image information from the most recent said first image information extracted by the first extracting means second extracting means The first image information extracted by the first extraction means and the first extraction means in order of the oldest second image information from the newest second image information extracted by the first extraction means. wherein the second image information and first image information representing an image including the extracted second image information and GaTeru engaged is by the same person are extracted.

  Further, in the present invention, the acquisition time by the image acquisition means of each of the first image information and the second image information extracted by the second extraction means is stored by the storage means. The storage means includes a RAM (Random Access Memory), an EEPROM (Electrically Erasable and Programmable Read Only Memory), a semiconductor storage element such as a flash EEPROM (Flash EEPROM), a floppy disk, and a CD-R (Compact Disc-Recordable). , A portable recording medium such as a CD-RW (Compact Disc-ReWritable), a magneto-optical disk, a fixed recording medium such as a hard disk, or an external storage device provided in a server computer connected to a network.

  That is, in the present invention, a region through which a person passes when entering and exiting a predetermined region is defined as a passing region, and the passing region is photographed every predetermined period, thereby allowing the human to pass the predetermined region. We are trying to capture the entry and exit with certainty.

  Therefore, in the image indicated by the first image information and the image indicated by the second image information extracted by the first extracting means, the same person has a high probability at a time interval separating the stay period in the predetermined area. Will be included.

  And in this invention, the person contained in each image shown by the above 1st image information and 2nd image information, ie, the person who enters into a predetermined area, and the person who leaves a predetermined area, are collated, and it is the same First image information and second image information indicating an image including a person are extracted by the second extracting means, and each of the extracted first and second image information acquisition times (each shooting time, entrance time) And the exit time) can be stored accurately, so that the entrance time and the exit time for the predetermined area of the same person can be accurately obtained, and as a result, the stay period of the person in the predetermined area can be accurately grasped. be able to.

As described above, according to the image processing apparatus of the first aspect, a plurality of pieces of image information are obtained by photographing a passing area through which a person passes when entering and exiting the predetermined area for each predetermined period. And comparing the acquired plurality of image information with the background image information predetermined as the image information acquired in a state where the person does not pass through the passage area . Since the moving direction of the difference area on the image in time series order enters the predetermined area, the person passes the passing area to enter the predetermined area with the image information substantially matching the passing direction when the person passes the passing area. Extracted as first image information acquired by shooting when passing, and image information that substantially matches the passing direction when a person passes through the passing region in order for the moving direction to leave the predetermined region. To get out of the area Most serial passage region extracted as the second image information obtained by photographing when the person is passing, in the direction toward the oldest first image information from the most recent first image information extracted, the extracted in order of the oldest second image information from the new second image information, the first image information and second image information indicating an image including the irradiation combined and the same person that is extracted from the first image information extracted And the second image information is extracted, and the respective acquisition times of the extracted first image information and second image information are stored, so that it is possible to accurately grasp the stay period of the person in the predetermined area, As a result, the movement status of the facility user can be grasped at a high level.

According to the present invention, as in the invention described in claim 2, the first extraction unit includes all the pixel data corresponding to each of the passage areas of the plurality of pieces of image information acquired by the image acquisition unit. The absolute value of the difference from the corresponding pixel data of the background image information is integrated for each of the plurality of image information, and the image information whose integration result is a predetermined value or more is extracted.

Further, as in the invention described in claim 3 , the first extraction unit compares each of the plurality of pieces of image information acquired by the image acquisition unit with the background image information, and compares the background image information with the background image information. It is preferable that a difference area whose difference area on the image is equal to or smaller than a predetermined width is an extraction target.
Further, as in a fourth aspect of the present invention, the first extraction unit is configured such that the luminance level of the image indicated by the plurality of pieces of image information acquired by the image acquisition unit is the luminance of the image indicated by the background image information. It is preferable to extract the first image information and the second image information after adjusting the plurality of pieces of image information so that the level is in accordance with the level. Thereby, the first image information and the second image information can be extracted with higher accuracy.

On the other hand, the second extraction means of the present invention, like the invention according to claim 5 , converts an image area of a person included in each image indicated by the first image information and the second image information into an upper body area and It is preferable to divide into lower body areas and extract the first image information and the second image information in which the degree of coincidence of both the upper body area and the lower body area is higher than a predetermined degree as indicating an image including the same person. Thereby, the first image information and the second image information indicating the image including the same person can be extracted with higher accuracy.
In particular, in this case, as in the invention described in claim 6 , the image acquisition means acquires the image information in color, and the second extraction means applies the image indicated by the first image information. The degree of coincidence of the distribution state between each color component of the upper body area of the included person and each color component of the upper body area of the person included in the image indicated by the second image information is higher than a predetermined degree, and the first image information When the degree of coincidence of the distribution state of each color component of the lower body area of the person included in the image indicated by the image and the color component of the lower body area of the person included in the image indicated by the second image information is higher than a predetermined degree It is preferable to extract the first image information and the second image information as indicating images including the same person.
Further, as in the invention described in claim 7 , the second extracting means is predetermined in a histogram of luminance values for each color component of the upper body region of the person included in the image indicated by the first image information. The amount of overlap between the luminance region range that is equal to or greater than the value and the luminance region range that is equal to or greater than the predetermined value in the histogram of luminance values for each color component of the upper body region of the person included in the image indicated by the second image information Indicated by the second image information and the luminance region range that is greater than or equal to a predetermined value in a histogram of luminance values for each color component of the lower body region of the person included in the image indicated by the first image information. Integrated value of the amount of overlap with the brightness area range that is equal to or greater than a predetermined value in the histogram of brightness values for each color component of the lower body area of the person included in the image It is preferable to extract both the first image information and the second image information in the case where the predetermined cumulative value or more as an indication of the image including the same person.
Further, as in the invention according to claim 8 , the second extraction means is included in a pattern of a person's clothes included in the image indicated by the first image information and the image indicated by the second image information. It is preferable to extract the first image information and the second image information whose degree of coincidence with the pattern of the person's clothes is higher than a predetermined degree as indicating an image including the same person.

In this case, as in the invention according to claim 9 , the second extraction means has a luminance level in the same image area of the image indicated by the first image information and the image indicated by the second image information. The extraction is preferably performed after adjusting at least one of the first image information and the second image information so as to be equal. Thereby, the first image information and the second image information indicating the image including the same person can be extracted with higher accuracy.

Further, according to the present invention, as in the invention described in claim 10 , the storage means stores in advance excluded feature amount information indicating a predetermined feature amount of an image of a person who is to be excluded from the grasping target of the entry / exit status. And storing the feature amount information indicating the feature amount of the person's image indicated by the first image information and the second image information against the excluded feature amount information, so as to match the first image information and the first image information. Determining means for determining whether the person indicated by the two-image information is the person to be excluded, and the first image information and the second image information determined by the determining means to be the person to be excluded. An exclusion means for excluding from the processing target can be provided.

  According to the present invention, the excluded feature amount information indicating the predetermined feature amount of the image of the person who is to be excluded from the entry / exit status grasp target is stored in advance by the storage unit. In addition, as the feature quantity, a physical quantity indicating the color of clothing (for example, a uniform of an employee working in a predetermined area of the present invention) worn by a person who wants to be excluded, or a pattern of the clothing. The physical quantity to show etc. can be illustrated.

  In the present invention, the determination means collates the feature amount information indicating the feature amount of the human image indicated by the first image information and the second image information with the excluded feature amount information, and the first image is checked. It is determined whether or not the person indicated by the information and the second image information is the person who wants to be excluded, and the first image information and the second image information that are determined by the determining means to be excluded by the determining means. Excluded from processing.

As described above, according to the present invention, the excluded feature amount information indicating the predetermined feature amount of the image of the person who is to be excluded from the entry / exit status grasp target is stored in advance, and the first image information and the second image information are stored. The feature amount information indicating the feature amount of the person's image indicated by the above and the excluded feature amount information are collated to determine whether the person indicated by the first image information and the second image information is a person to be excluded. Since the first image information and the second image information that are determined to be excluded are excluded from the processing target, only the person who wants to grasp the entry / exit status with respect to the predetermined area can be targeted. The usefulness of can be improved.
Particularly, in this case, as in the invention described in claim 11 , the storage unit stores in advance the clothing feature information indicating the feature of the clothing of the person desired to be excluded as the excluded feature amount information, and the determination unit includes: The first image information and the second image information indicate the characteristics of the person's clothing indicated by the first image information and the second image information and the characteristics of the clothing indicated by the clothing characteristic information. It is preferable to determine whether or not a person is the person who wants to be excluded.

  By the way, since the extraction by the first extraction means and the second extraction means in the present invention is performed automatically, there may be a case where an extraction omission occurs.

Therefore, as in the invention described in claim 12 , the present invention provides each image indicated by the first image information and the second image information that has not been extracted by the first extraction means, and the second extraction means. Display means for displaying each image indicated by the first image information and the second image information not extracted by the display means, and in the image displayed by the display means, the passage area for entering the predetermined area A first designation for designating an image when a person is passing through and a second designation for designating an image when a person is passing through the passing area to exit the predetermined area. A designation instruction for inputting, a designation instruction for associating the image designated by the first designation and the image designated by the second designation, each including the same person. The first image information indicating the image specified by the first specification and the image specified by the second specification, and the image specified by the second specification, which are associated by the association instruction input by the association instruction input unit. It is preferable to further comprise control means for performing control so that the acquisition time of each of the second image information by the image acquisition means is stored in the storage means.

  According to the present invention, each image indicated by the first image information and the second image information not extracted by the first extraction means, and the first image information and the second image information not extracted by the second extraction means. Each image shown is displayed by the display means. At this time, the image displayed by the display means may be an image indicated by all image information not extracted by the first extraction means and the second extraction means, or only an image indicated by a part of the image information. It is good. The display means includes a display such as a liquid crystal display, a CRT display, a plasma display, and an organic EL display.

In the present invention, in the image displayed by the display means, a first designation for designating an image when a person is passing through the passing area to enter the predetermined area, and the above-mentioned for exiting from the predetermined area. At least one of the specified second specified is inputted by the specification input means for designating an image when the passage region people are passing through.
In the present invention, the association instruction input unit inputs an association instruction for associating the image designated by the first designation including the same person with the image designated by the second designation.
The designation input unit and the association instruction input unit include a pointing device such as a mouse, a track ball, and a tablet operated to input the first designation, the second designation, and the association instruction , and a keyboard. Various switches are included.

That is, in the present invention, the display means displays the image information that has not been extracted by the first extraction means and the second extraction means of the present invention, that is, the image based on the image information that may have been leaked from the extraction target by each extraction means. while, after designating the image information being subject to extraction by the first extraction means from among these, has been to specify the image information which is the extraction target by the second extraction means, whereby Image information leaked from the extraction target by the first and second extraction means can be extracted by visual judgment by a person with respect to the image displayed on the display means.

Thus, according to the present invention, each image indicated by the first image information and the second image information that has not been extracted by the first extraction means, and the first image information and the first image information that have not been extracted by the second extraction means. Each image indicated by the two-image information is displayed on the display means, and in the displayed image, a first designation for designating an image when a person passes through the passage area to enter the predetermined area, the predetermined designation A second designation for designating an image when a person is passing through the passing area to exit the area, and an image designated by the first designation and the image designated by the second designation, each including the same person. Since the association instruction to be associated can be input, the extraction omission by the first extraction unit and the second extraction unit of the present invention can be compensated.

In order to achieve the above object, an image processing program according to claim 13 is an image processing program for performing image processing for grasping a person's entering / exiting status with respect to a predetermined area, when entering the predetermined area; An image acquisition step of acquiring a plurality of pieces of image information by photographing a passage region through which a person passes when exiting the predetermined region, and a plurality of pieces of image information acquired by the image acquisition step and the passage region. Compared with the background image information predetermined as the image information acquired by the image acquisition step in a state where no person has passed, and from the plurality of image information, the difference area on the image with the background image information before to enter the image information substantially coincides with the passage direction when the moving direction in chronological order to the passage region human pass to enter the predetermined region in the predetermined region Substantially with the passage region human is extracted as first image information obtained by photographing it passes through, in the passing direction when the moving direction is the passage region human pass to exit from the predetermined area A first extraction step of extracting the matching image information as second image information acquired by photographing when a person is passing through the passage region in order to exit the predetermined region, and extraction by the first extraction step In the direction from the newest first image information to the oldest first image information, in the order of the oldest second image information from the newest second image information extracted by the first extraction step, It includes irradiation combined and the same person and the extracted second image information by the the extracted first image information first extraction step by the first extraction step A second extraction step for extracting the first image information indicating the image and the second image information; and the image acquisition step for each of the first image information and the second image information extracted by the second extraction step. And a storage step of storing the acquisition time in the storage means.

Therefore, according to the image processing program of the thirteenth aspect , the computer can be operated in the same manner as the first aspect of the invention. As a result, it is possible to grasp the movement status of the facility user at a high level.

According to the present invention, a plurality of pieces of acquired image information and passages are acquired by photographing a passing region through which a person passes when entering a predetermined region and exiting from the predetermined region for each predetermined period. Compared with background image information predetermined as image information acquired in a state where no person has passed through the region, from a plurality of image information in the time-series order of the difference region on the image with the background image information Image information that substantially matches the passing direction when a person passes through the passing area because the moving direction enters the predetermined area was acquired by photographing when the person was passing the passing area to enter the predetermined area. The first image information is extracted, and the person passes through the passage area in order to exit the predetermined area of image information that substantially matches the passage direction when the person passes through the passage area in order to move out of the predetermined area. When passing Extracted as second image information obtained by photographing, in the direction towards the oldest first image information from the most recent first image information extracted, the oldest second image information from the most recent second image information extracted successively with a, and a second image information extracted from the first image information extracted irradiation combined extracts said second image data and the first image information representing an image including the same person, extracted Since the acquisition time of each of the first image information and the second image information is stored, it is possible to accurately grasp the staying period of the person in the predetermined area, and as a result, the movement situation of the facility user is high. The effect of being able to grasp by level is obtained.

  The best mode for carrying out the present invention will be described below in detail with reference to the drawings. In addition, the form at the time of applying this invention to the use which grasps | ascertains the movement state of the patient with respect to the examination room of a hospital is demonstrated here.

[First Embodiment]
First, the configuration of a facility user measurement system 10 to which the present invention is applied will be described with reference to FIG.

  As shown in the figure, the facility user measurement system 10 according to this embodiment includes a facility user measurement device 20. The facility user measuring device 20 includes a control unit 36 for controlling the overall operation of the apparatus, a keyboard 30 and a mouse 32 used for inputting various information from the operator, processing results and various menus by the apparatus. A display 34 for displaying a screen, a message, and the like is provided. That is, the facility user measuring device 20 according to the present embodiment is configured by a general-purpose personal computer.

  On the other hand, the facility user measurement system 10 includes a digital electronic still camera (hereinafter referred to as “camera”) 50 that captures a color still image for each preset period. It is electrically connected to the person measuring device 20. In addition, although there is no limitation in the connection form of the camera 50 and the facility user measuring device 20, in this Embodiment, it connects via the network. Thereby, even if the camera 50 and the facility user measuring device 20 are relatively distant from each other, they can be easily connected, and the facility user measuring system 10 can be flexibly constructed.

  Next, the configuration of the electrical system of the facility user measurement system 10 will be described with reference to FIG.

  As shown in the figure, the facility user measuring device 20 is used as a CPU (central processing unit) 22 that controls the operation of the facility user measuring device 20 as a whole and a work area when the CPU 22 executes various processing programs. A RAM (Random Access Memory) 24, a ROM (Read Only Memory) 26 in which various processing programs and various parameters are stored in advance, a hard disk 28 used for storing various information, the keyboard 30 and the mouse 32 described above. , And the display 34 and an external interface 36 that controls transmission and reception of various information between the devices connected to the outside are connected to each other via a system bus BUS. The external interface 36 is connected to an output terminal for outputting digital image data indicating a color still image obtained by photographing of the camera 50 described above.

  Therefore, the CPU 22 accesses the RAM 24, the ROM 26, and the hard disk 28, acquires various input information via the keyboard 30 and the mouse 32, displays various information on the display 34, and the digital image from the camera 50 via the external interface 36. Data acquisition can be performed.

  The camera 50 is set to output the digital image data obtained by photographing to the facility user measuring apparatus 20 after compressing the digital image data in a predetermined compression format (here, JPEG format). Realizes high-speed data transfer.

  In addition, as shown in FIG. 3, the camera 50 can take an image of the entrance / exit 66 of the examination room 62, which is the subject of grasping the patient's movement status, in the vicinity of the boundary between the ceiling surface and the wall surface in the waiting room 64 of the hospital 60. In place.

  FIG. 4 shows an example of an image 70 obtained by photographing with the camera 50. In the figure, for convenience, the image 70 is represented by a line drawing, but it goes without saying that the image 70 is actually a color image having a predetermined resolution.

  As shown in the figure, this image 70 includes the entrance / exit 66 of the examination room 62 and the inside of the waiting room 64.

  By the way, in the facility user measurement system 10 according to the present embodiment, as shown in FIG. 5, a rectangular area having a predetermined area including the entrance / exit port 66 in the shooting area by the camera 50 is predetermined as the determination area A1. ing. Further, in the facility user measurement system 10, a rectangular area corresponding to a wall surface with a very low possibility that a person in the photographing area is present is predetermined as the luminance correction area A2. Note that the positions of the determination area A1 and the luminance correction area A2 shown in the figure are examples, and can be changed as appropriate.

  In the facility user measurement system 10 according to the present embodiment, the digital image data corresponding to the determination area A1 of the digital image data obtained by photographing with the camera 50 when no person is present in the determination area A1 is used as the background image. Information is stored in advance in a predetermined area of the hard disk 28 of the facility user measuring apparatus 20. Similarly, in the facility user measurement system 10, the digital image data corresponding to the luminance correction area A2 of the digital image data obtained by photographing with the camera 50 when no person is present in the luminance correction area A2 is used as the luminance correction image. Information is stored in advance in a predetermined area of the hard disk 28.

  The facility user measurement system 10 according to the present embodiment has a function of discriminating employees and patients such as doctors, nurses, and radiographers working at the hospital 60 based on the characteristics of clothing. Yes. In a predetermined area of the hard disk 28, information indicating the color component (hereinafter referred to as “excluded color component”) of the clothes of the employee (in this embodiment, white robe) is stored in advance.

  Next, the operation of the facility user measurement system 10 according to the present embodiment will be described. First, referring to FIG. 6, the facility user that is executed within a predetermined measurement period (in this embodiment, a period from 9:00 am to 4:30 pm) in facility user measurement device 20. The measurement process will be described. FIG. 6 is a flowchart showing the processing flow of the facility user measurement processing program executed by the CPU 22 of the facility user measuring device 20 during the measurement period, and the program is stored in a predetermined area of the ROM 26 in advance. Yes. Here, a case will be described in which a shooting interval (here, 0.2 seconds) by the camera 50 is set in advance. That is, in the present embodiment, it is set so that digital image data indicating five color still images per second is transmitted from the camera 50 to the facility user measuring apparatus 20.

  First, in step 100, reception of digital image data for one color still image from the camera 50 is waited for, and in the next step 102, the received digital image data is stored in a predetermined area of the hard disk 28. At this time, the digital image data is stored as an electronic file (image file), and information (so-called time stamp) indicating the time when the digital image data is acquired is automatically stored in the image file. Will be.

  In the next step 104, it is determined whether or not the digital image data received from the camera 50 and stored in the hard disk 28 has reached a predetermined number (in this embodiment, the number of digital image data obtained by photographing for 2 seconds). If the determination is negative and the determination is negative, the process returns to step 100. When the determination is affirmative, the process proceeds to step 106.

  The repetitive processing from step 100 to step 104 is for performing the processing after step 106 every time the predetermined number of digital image data is stored. The number of digital image data that has not been processed in step 106 and subsequent steps is compared with the predetermined number. In the following description, the predetermined number of digital image data to be processed is referred to as “processing target image data”. In the present embodiment, digital image data indicating 10 images corresponding to the number of digital image data obtained by shooting for 2 seconds is image data to be processed.

  In step 106, the processing target image data is adjusted so that the luminance level in the determination area A1 of the image indicated by the processing target image data becomes a level corresponding to the luminance level of the image indicated by the background image information. The adjustment is performed after each of the predetermined number of pieces of processing target image data is expanded in a format corresponding to the compression format.

  In the present embodiment, the luminance level is adjusted as follows as an example.

  First, the pixel data R (red), G (green) of the background image information corresponding to a partial region (for example, a region near the upper end of the determination area A1) that is unlikely to pass through in the determination area A1. The average values BR, BG, and BB for each B (blue) are calculated, and the average values TR, TG, and TB for the pixel data of each processing target image data corresponding to the partial area are calculated for each R, G, and B. To do.

  Next, a coefficient CR (= BR / TR) indicating a ratio of the average value TR to the average value BR, a coefficient CG (= BG / TG) indicating a ratio of the average value TG to the average value BG, and an average with respect to the average value BB A coefficient CB (= BB / TB) indicating the ratio of the value TB is calculated for each processing target image data.

  Finally, for each piece of processing target image data, a coefficient CR in the processing target image data for each pixel data of R in the processing target image data determination area A1, and each of G in the processing target image data determination area A1. The pixel data is multiplied by the coefficient CG in the processing target image data, and the pixel data of B in the processing target image data determination area A1 is multiplied by the coefficient CB in the processing target image data.

  Note that the method of adjusting the luminance level of the processing target image data is not limited to the above method. For example, when the digital image data obtained by the camera 50 has a format indicated by a luminance signal and a color difference signal, the luminance signal is applied instead of the R, G, and B values to be processed. The brightness level in the image data determination area A1 can be adjusted. Further, for example, the coefficient is not calculated for all of R, G, and B as described above, but the coefficient is calculated for only one color component of R, G, and B, and the coefficient is calculated for all the color components. It can also be applied to the adjustment of the luminance level.

  When the adjustment of the luminance level in the determination area A1 of the processing target image data is completed, in the next step 108, a difference value D between the processing target image data and the background image information is calculated for each processing target image data.

  In the present embodiment, the difference value D is calculated as follows as an example.

  First, the absolute value of the difference between all the pixel data in the determination area A1 of the processing target image data and the corresponding pixel data of the background image information is calculated for each processing target image data for each of R, G, and B. .

  Next, the difference value D is calculated by integrating all the calculated absolute values for each processing target image data.

  Note that the method of calculating the difference value D is not limited to the above method. For example, when the digital image data obtained by the camera 50 has a format indicated by a luminance signal and a color difference signal, at least one of the luminance signal and the color difference signal is used instead of the R, G, and B values. By applying, the difference value D between the processing target image data and the background image information can be calculated.

  Note that, at the time of calculating the difference value D, for each processing target image data, an area changed from the image indicated by the background image information of the image indicated by the processing target image data (hereinafter referred to as “image change area”). .) Is extracted, and information indicating the image change area is stored in a predetermined area of the hard disk 28 in association with the corresponding processing target image data.

  Here, in the present embodiment, the extraction of the image change area is performed by extracting the absolute value of at least one of R, G, and B with a predetermined value (for example, the maximum value of pixel data (pixel data has an 8-bit configuration). In some cases, the value is half of “255”).) The above pixel group is extracted as the image change region. However, as another method, for example, the absolute value of all of R, G, and B is used. Any area that can extract a change area (difference area) between image data to be processed and background image information, such as a mode in which a pixel group having a value equal to or greater than the predetermined value is extracted as the image change area The method can be applied.

  When the calculation of the difference value D for each processing target image data is completed, in the next step 110, it is determined whether or not there is a difference value D that is equal to or greater than a predetermined value, and a negative determination is made. In this case, it is determined that there is no person image included in the processing target image data, and the process returns to step 100. If the determination is positive, the process proceeds to step 112. The predetermined value applied here may include a human image in the determination area A1 of the image indicated by the processing target image data corresponding to the difference value D when the difference value D is equal to or greater than the value. As a value that can be regarded as having high performance, a value statistically obtained in advance by experiment, a value obtained in advance by computer simulation, or the like can be applied.

  In step 112, the width W of the image change area in all the processing target image data corresponding to the difference value D for which affirmative determination is made in step 110 is calculated. In the next step 114, the calculated width W is a predetermined width ( Here, it is determined whether or not there is processing target image data that is equal to or smaller than a width of a standard person, and in the case of a negative determination, only one person is included in the processing target image data. It is determined that there is nothing to be performed, and the process returns to step 100. If the determination is affirmative, the process proceeds to step. Here, when the process proceeds to step 116, the processing target image data whose width W is equal to or smaller than the predetermined width is processed into image data (hereinafter, referred to as a target for subsequent processing) by processing such as setting a predetermined flag. It can be identified as being “actual processing target image data”).

  In step 116, the degree of coincidence C between the information indicating the color component in the image change area and the information indicating the excluded color component is calculated for each actual processing target image data. Here, the ratio (%) of the number of pixels that are substantially the same color component as the excluded color component to the total number of pixels in the image change area is the matching degree C, but the present invention is not limited to this.

  In the next step 118, it is determined for each actual processing target image data whether or not the calculated degree of coincidence C is equal to or greater than a predetermined value (in this case, 90%). On the other hand, an attribute indicating that the employee is an employee is assigned in step 120, and an attribute indicating that the employee is a non-employee is assigned in step 122 to the actual processing target image data that is negatively determined. Migrate to

  In step 124, the change direction of the image change area in the image indicated by the actual processing target image data is detected. In the present embodiment, as shown in FIG. 7 as an example, the movement direction of the person when entering the examination room 62 in the determination area A1 is a direction from left to right (hereinafter referred to as “room entry direction”). Since the direction of movement of the person when leaving the examination room 62 is the direction from right to left (hereinafter referred to as “leaving direction”), one of the entering direction and the leaving direction is detected here. Will be.

  In the present embodiment, when there are a plurality of actual process target image data, the change direction in the time series order of the image change area indicated by each actual process target image data is detected as the change direction. In addition, when there is only one actual processing target image data, when the position in the determination area A1 of the image change area indicated by the actual processing target image data is closer to the left end, it is assumed that the direction is the entrance direction. When it is closer to the right end, it is assumed that the direction is to leave the room.

  In the next step 126, it is determined for each actual process target image data whether or not the detected change direction is the entrance direction. An attribute indicating that the image is a captured image is assigned, and an attribute indicating that the image is a captured image at the time of leaving the room is assigned to the actual processing target image data that is negatively determined, and then the process proceeds to step 132.

  In step 132, information (hereinafter referred to as “measurement information”) regarding the actual processing target image data obtained by the above processing is stored in a predetermined area of the hard disk 28. Here, when there are a plurality of pieces of actual processing target image data, only information related to the actual processing target image data having the largest proportion of the image change area is stored.

  In the next step 134, it is determined whether or not a predetermined measurement period (in this embodiment, the period from 9 am to 4:30 pm) has ended. Returning to Step 100, the facility user measurement processing program is terminated when an affirmative determination is made.

  FIG. 8 shows an example of measurement information stored in the hard disk 28 by the facility user measurement processing program. As shown in the figure, in the facility user measurement system 10 according to the present embodiment, all the digital image data stored in the hard disk 28 is given as the measurement information in the order of time series stored in the hard disk 28. An event number, an image file name, a shooting date and time (here, a time stamp given to the image file), an entry / exit attribute indicating whether the entry or exit is given in step 128 or 130, and Each information of the employee attribute indicating whether the employee or the non-employee is given in step 120 or 122 is stored in the hard disk 28 for each actual processing target image data. In the figure, “IN” indicates entry and “OUT” indicates exit. In the figure, 'P' indicates a non-employee (patient) and 'S' indicates an employee. Below, the case where measurement information is what is shown in FIG. 8 is demonstrated.

  Next, with reference to FIG. 9, the individual identification process performed after completion | finish of execution of the said facility user measurement process in the facility user measurement apparatus 20 is demonstrated. FIG. 9 is a flowchart showing the flow of processing of the individual identification processing program executed by the CPU 22 of the facility user measuring device 20 when the individual identification processing is executed. The program is also stored in a predetermined area of the ROM 26 in advance. ing.

  First, in step 200, the measurement information (see also FIG. 8) stored by the facility user measurement process is read from the hard disk 28, and the actual processing target image data corresponding to the image file name recorded in the measurement information is read. Classification based on the attribute information included in the measurement information for each type of person (here, patient and employee) and entry / exit, and information indicating the classification result (hereinafter “Classification”) Information ”) is stored in a predetermined area of the hard disk 28.

  FIG. 10 shows an example of the classification information obtained by the processing of step 200. As shown in the figure, in the facility user measurement system 10 according to the present embodiment, as the classification information, the image file name recorded in the measurement information is classified according to the type of the person being photographed, and enters / exits the room. The data is classified into four different classification areas and stored in the hard disk 28. As shown in the figure, at this time, the image file names are stored in each classification area so that the image file names are arranged in chronological order of the shooting date and time. Below, the case where classification information is what is shown in FIG. 10 is demonstrated. In the following, the classification area of the actual processing target image data at the time of entering the room is called “entry image data area”, and the classification of the actual processing target image data at the time of leaving the patient is shown. The area is referred to as “exit image data area”.

  In the next step 202, the image file name with the latest shooting date and time in the room image data area that has not been subjected to subsequent processing is read from the classification information, and the digital image data of the image file (hereinafter referred to as “process target room image data”). Is read from the hard disk 28.

  In the next step 204, the luminance level in the luminance correction area A2 of the image indicated by the processing target entry image data is set to a level corresponding to the luminance level of the image indicated by the luminance correction image information stored in the hard disk 28 in advance. The processing target entry image data is adjusted. The adjustment is performed after the processing target entry image data is expanded in a format corresponding to the compression format.

  In the present embodiment, the luminance level is adjusted as follows as an example.

  First, the average values MR, MG, and MB for each pixel data of the luminance correction image information for each of R, G, and B are calculated, and the R, G, and B of each pixel data in the luminance correction area A2 of the processing target room entrance image data are calculated. Average values NR, NG, and NB are calculated for each.

  Next, a coefficient DR (= MR / NR) indicating the ratio of the average value NR to the average value MR, a coefficient DG (= MG / NG) indicating the ratio of the average value NG to the average value MG, and an average for the average value MB A coefficient DB (= MB / NB) indicating the ratio of the value NB is calculated.

  Finally, the coefficient DR is multiplied for each R pixel data in the processing target room entrance image data, the coefficient DG is multiplied for each G pixel data, and the coefficient DB is multiplied for each B pixel data.

  Note that the method for adjusting the luminance level of the processing target entry image data is not limited to the above method. For example, when the digital image data obtained by the camera 50 has a format indicated by a luminance signal and a color difference signal, the luminance signal is applied instead of the R, G, and B values to be processed. The brightness level of the entrance image data can be adjusted. Further, for example, the coefficient is not calculated for all of R, G, and B as described above, but the coefficient is calculated for only one color component of R, G, and B, and the coefficient is calculated for all the color components. It can also be applied to the adjustment of the luminance level.

  In the next step 206, predetermined feature data is extracted from the processing target entry image data after the luminance level correction. In the facility user measurement system 10 according to the present embodiment, the feature data is extracted as follows.

  First, information indicating the image change area stored in the hard disk 28 in the processing of step 108 of the facility user measurement processing program for the image indicated by the processing target entry image data is read.

  Next, assuming that the image change area indicated by the read information is an area indicating a person, as shown in FIG. 11, the area includes a head area, an upper body area, a lower body area, Divide into Note that here, the division is performed according to the ratio of each area of the head, upper body, and lower body in a standard person, but is not limited thereto, for example, each of the head, upper body, and lower body It is good also as what divides | segments according to the shape of an area | region.

  Next, for each region of the upper body and the lower body, a histogram of luminance values for each color component in the processing target entry image data is derived as the feature data.

  Note that the present invention is not limited to this derivation method, and for example, an average value of luminance values for each region of the upper body and lower body and for each color component may be derived as the feature data.

  In the next step 208, the latest image file name with the latest shooting date and time in the leaving image data area that has not been subjected to the subsequent processing is read from the classification information (see also FIG. 10), and the digital image data (hereinafter referred to as the image file) of the image file is read. , “Processing target leaving image data”) is read from the hard disk 28.

  In the next step 210, the brightness level in the brightness correction area A2 of the image indicated by the processing target leaving image data becomes a level corresponding to the brightness level of the image indicated by the brightness correction image information stored in the hard disk 28 in advance. The processing target leaving image data is adjusted. The adjustment is performed after the processing target leaving image data is expanded in a format corresponding to the compression format.

  In the present embodiment, the luminance level is adjusted as follows as an example.

  First, average values MR, MG, MB for each R, G, B of each pixel data of the brightness correction image information are calculated, and R, G, B of each pixel data in the brightness correction area A2 of the processing target leaving image data. Average values OR, OG, and OB are calculated for each.

  Next, a coefficient ER (= MR / OR) indicating a ratio of the average value OR to the average value MR, a coefficient EG (= MG / OG) indicating a ratio of the average value OG to the average value MG, and an average with respect to the average value MB A coefficient EB (= MB / OB) indicating the ratio of the value OB is calculated.

  Finally, the coefficient ER is multiplied by the R pixel data in the processing target leaving image data, the coefficient EG is multiplied by the G pixel data, and the coefficient EB is multiplied by the B pixel data.

  Note that the method of adjusting the luminance level of the processing target leaving image data is not limited to the above method. For example, when the digital image data obtained by the camera 50 has a format indicated by a luminance signal and a color difference signal, the luminance signal is applied instead of the R, G, and B values to be processed. The brightness level of the leaving image data can be adjusted. Further, for example, the coefficient is not calculated for all of R, G, and B as described above, but the coefficient is calculated for only one color component of R, G, and B, and the coefficient is calculated for all the color components. It can also be applied to the adjustment of the luminance level.

  In the next step 212, predetermined feature data is extracted from the processing target leaving image data after correcting the luminance level. In the facility user measurement system 10 according to the present embodiment, the feature data is extracted as follows.

  First, information indicating the image change area stored in the hard disk 28 in the processing of step 108 of the facility user measurement processing program for the image indicated by the processing target leaving image data is read.

  Next, assuming that the image change area indicated by the read information is an area indicating a person, as shown in FIG. 11, the area includes a head area, an upper body area, a lower body area, Divide into Note that here, the division is performed according to the ratio of each area of the head, upper body, and lower body in a standard person, but is not limited thereto, for example, each of the head, upper body, and lower body It is good also as what divides | segments according to the shape of an area | region.

  Next, for each region of the upper body and the lower body, a histogram of luminance values for each color component in the processing target leaving image data is derived as the feature data.

  Note that the present invention is not limited to this derivation method, and for example, an average value of luminance values for each region of the upper body and lower body and for each color component may be derived as the feature data.

  In the next step 214, the degree of coincidence JC between the upper body feature data extracted in step 206 and the upper body feature data extracted in step 212, the lower body feature data extracted in step 206, and the above step 212. The degree of coincidence KC with the feature data of the lower body extracted in step S is calculated.

  As shown in FIG. 11, in the facility user measurement system 10 according to the present embodiment, the luminance region range that is equal to or greater than a predetermined threshold in the histogram derived as the upper body feature data in step 206 as the degree of coincidence JC. And an integrated value of the overlap amount for each color component of the luminance region range that is equal to or greater than the predetermined threshold in the histogram derived as upper body feature data in step 212 is derived.

  Further, the facility user measurement system 10 derives the degree of coincidence KC as a luminance region range that is equal to or greater than a predetermined threshold in the histogram derived as the lower body feature data in step 206 and the lower body feature data in step 212. An integrated value of the overlap amount for each color component of the luminance region range that is equal to or greater than the predetermined threshold in the histogram is derived.

  The method of calculating the degree of coincidence JC and the degree of coincidence KC is not limited to the above method. For example, the degree of coincidence JC and the degree of coincidence KC may be derived by pattern matching for each color component with respect to the shape of each histogram derived as feature data of the upper body and lower body. it can.

  In the next step 216, it is determined whether or not both the degree of coincidence JC and the degree of coincidence KC derived in step 214 are equal to or greater than a predetermined value. If the determination is affirmative, the image indicated by the processing target entry image data And the person included in the image indicated by the processing target leaving image data is determined to be the same person, and the process proceeds to step 218 to associate the processing target entering image data with the processing target leaving image data. After information (hereinafter referred to as “corresponding information”) is recorded, the process proceeds to step 226. The predetermined value applied in step 216 is a person who wears the same clothing on the two processing target image data corresponding to the degree of coincidence when the degree of coincidence JC and the degree of coincidence KC are equal to or higher than the values. As a value that can be considered to be highly likely to include the image, a value statistically obtained in advance by an experiment, a value obtained in advance by a computer simulation, or the like can be applied.

  FIG. 12 shows an example of association information according to the present embodiment. As shown in the figure, here, the event number of the processing target entry image data to be associated, the entry time indicated by the shooting date and time of the processing target entry image data, the event number of the processing target exit image data to be associated, the processing target exit image Each information of the exit time indicated by the shooting date and time of the data and the stay period obtained as the difference between the entrance time and the exit time is recorded as association information.

  On the other hand, if a negative determination is made in step 216, the process proceeds to step 220, where it is determined whether only one of the coincidence degree JC and the coincidence degree KC is equal to or greater than the predetermined value. If so, go to Step 222.

  In step 222, the weighting coincidence WC is calculated by the following equation (1).

WC = JC × WJ + KC × WK (1)
Here, WJ represents a weight value for the degree of coincidence JC in the upper body, and WK represents a weight value for the degree of coincidence KC in the lower body.

  A patient who enters the examination room may leave the room with his / her upper coat taken off in the examination room. In this case, the degree of coincidence JC is not appropriate as a value for specifying the same person. On the other hand, since there are few patients who leave the room with their clothes on the lower body, the coincidence degree KC is more reliable as a value for identifying the same person than the coincidence degree JC.

  Considering the above points, in the facility user measurement system 10 according to the present embodiment, the weight value WK is set to a value larger than the weight value WJ, and thereby, a higher degree of coincidence is derived. I am doing so. However, the present invention is not limited to this. For example, when the degree of coincidence WC is simply derived, only the degree of coincidence KC is directly applied as the degree of coincidence WC, or only by adding the degree of coincidence JC and the degree of coincidence KC without weighting. The degree of coincidence WC can also be derived.

  In the next step 224, the derived weighted coincidence degree WC is stored in a predetermined area of the hard disk 28 in association with the processing target entry image data and the processing target exit image data, and then the process proceeds to step 226.

  In Step 226, by setting a processed flag (not shown) corresponding to the processing target entry image data and the processing target leaving image data, it is recorded that these image data have been processed, and then the process proceeds to Step 228. It is determined whether or not the above processing is completed for the digital image data of all image file names stored in the entry image data area of the classification information (see also FIG. 10). Returning to step 202, when the determination is affirmative, the individual recognition processing program is terminated.

  On the other hand, if a negative determination is made in step 220, the process proceeds to step 230, in which it is determined whether or not the processing target leaving image data is the oldest stored in the leaving image data area of the classification information. When the determination is made, the process returns to step 202. When the determination is negative, the process proceeds to step 232, and a new image file name of the next shooting date and time stored in the leaving image data area of the classification information is entered. Read from the classification information, read the digital image data of the image file name from the hard disk 28 as the processing target leaving image data, and then return to step 210.

  With this individual recognition processing program, the target entry image data is processed in the order from the newest processing target exit image data to the oldest processing target exit image data in the direction from the newest processing target entrance image data to the oldest processing target entrance image data. And the processing target leaving image data are collated, and when image data having a high degree of coincidence is detected by the collation, the image data is associated and excluded from the subsequent collation targets.

  Usually, the last person entering the examination room has a high possibility of matching with the person who left the room, and therefore, high-speed processing is realized by using the above collation procedure.

  As described above in detail, in the present embodiment, the determination area A1 through which a person passes when entering the examination room 62 and exiting the examination room 62 is set for a predetermined period (in this embodiment, 0.2 seconds). ) To obtain a plurality of digital image data, and from the plurality of digital image data, when entering a room obtained by photographing when a person passes through the determination area A1 to enter the examination room 62 The digital image data and the digital image data at the time of leaving obtained by photographing when a person passes through the judgment area A1 to exit the examination room 62 are extracted and included in the image indicated by the digital image data at the time of entering the room Digital image data indicating images containing the same person is extracted by comparing the person included in the image indicated by the digital image data at the time of leaving and Since each acquisition time (entry time and leaving time) of the image data is stored in association with each other, it is possible to accurately grasp the stay period of the person in the examination room 62. As a result, the facility user The movement status can be grasped at a high level.

  In the present embodiment, digital image data acquired by the camera 50 in a state where no person passes through the determination area A1 is stored in advance as background image information, and a plurality of digital images acquired by the camera 50 is stored. By comparing the data with the background image information, the digital image data acquired by photographing when a person is passing through the determination area A1 is extracted. Therefore, the digital image data can be easily extracted. .

  In the present embodiment, the plurality of digital images are set such that the brightness level of the image indicated by the plurality of digital image data acquired by the camera 50 is a level corresponding to the brightness level of the image indicated by the background image information. Since the image data is adjusted, it is possible to extract digital image data acquired by photographing when a person passes through the determination area A1 with higher accuracy than in the case where the adjustment is not performed. .

  In the present embodiment, a plurality of digital image data acquired by the camera 50 is compared with the background image information, and a difference area on the image with the background image information (in this embodiment, an image change area). ) In order of movement in time-series order to enter the examination room 62, digital image data that substantially matches the passage direction when a person passes through the determination area A1 is extracted as digital image data at the time of entry, and the movement direction Is extracted as digital image data at the time of leaving the room because the digital image data that substantially matches the passing direction when a person passes through the judgment area A1 in order to exit the examination room 62 is obtained. Can be extracted.

  Further, in the present embodiment, the human image area included in each image indicated by the processing target entry image data and the processing target exit image data is divided into an upper body area and a lower body area, and both the upper body area and the lower body area are divided. Since the processing target entry image data and processing target exit image data whose degree of coincidence is higher than a predetermined degree are extracted as indicating images containing the same person, it is compared when collating without dividing into upper body area and lower body area Thus, image data indicating an image including the same person can be extracted with higher accuracy.

  Further, in the present embodiment, the luminance level in the same image area (in the present embodiment, the luminance correction area A2) of the image indicated by the processing target entry image data and the image indicated by the processing target leaving image data is equal. Since each image data is adjusted, image data indicating an image including the same person can be extracted with higher accuracy than in the case where the adjustment is not performed.

  Further, in the present embodiment, excluded feature amount information (excluded in the present embodiment) that indicates a predetermined feature amount of an image of a person (employee in the present embodiment) that is desired to be excluded from the grasp status of the entry / exit status. Color component) is stored in advance, the feature amount information indicating the feature amount of the person image indicated by the actual processing target image data is compared with the excluded feature amount information, and the person is to be excluded Since the actual processing target image data determined to be excluded is excluded from the processing target, only the person who wants to grasp the entrance / exit status of the examination room 62 (patient in the present embodiment) The utility of the facility user measurement system 10 can be improved.

  Furthermore, in this embodiment, digital image data is acquired in color. Therefore, when digital image data indicating an image including a person or digital image data indicating an image including the same person is extracted, color image data is extracted. Color components (R (red), G (green), B (blue), luminance signal and color difference signal) can be used, and each extraction accuracy is higher than when acquiring monochrome image information. Can be improved.

[Second Embodiment]
In the facility user measurement process (see FIG. 6) in the first embodiment, a person who enters the examination room 62 and a person who leaves the examination room 62 are detected by comparing the difference value D with a predetermined value. Therefore, it is not always possible to detect all persons entering and leaving the room. Similarly, in the individual recognition process (see FIG. 9) in the first embodiment, the same person is detected by comparing the coincidence degree JC and the coincidence degree KC with a predetermined value. It is not always possible to detect the same person. However, it goes without saying that detecting as many detection targets as possible is preferable for grasping the movement status of the patient.

  Therefore, in the second embodiment, a description will be given of a case in which what cannot be automatically detected by the facility user measurement process and the individual recognition process is detected as much as possible by interposing a person. Since the configuration of the facility user measurement system 10 according to the second embodiment is the same as that of the first embodiment, description thereof is omitted, and hereinafter, the second embodiment. The operation of the facility user measurement system 10 according to will be described.

  First, the facility user measurement process executed in the facility user measurement device 20 according to the second embodiment will be described. In the facility user measuring device 20 according to the present embodiment, processing substantially similar to the facility user measurement processing program shown in FIG. 6 is executed, but the difference value D calculated in step 108 in the figure and step 112 The only difference from the first embodiment is that each value of the width W derived in is stored in a predetermined area of the hard disk 28 in a state associated with the corresponding digital image data.

  Therefore, after execution of the facility user measurement processing program according to the second embodiment, the difference value D for all digital image data and the digital image data for which the difference value D is equal to or greater than a predetermined value. The width W is stored in the hard disk 28.

  After the facility user measurement process is executed, the facility user measurement apparatus 20 according to the second embodiment performs a measurement correction process. Hereinafter, the measurement correction process will be described with reference to FIG. FIG. 13 is a flowchart showing a processing flow of a measurement correction processing program executed by the CPU 22 of the facility user measuring apparatus 20 when the measurement correction processing is executed. The program is also stored in a predetermined area of the ROM 26 in advance. ing.

  First, in step 300, all the difference values D and widths W stored by the facility user measurement process are read from the hard disk 28.

  In the next step 302 to step 308, by detecting digital image data in which the difference value D is greater than or equal to a predetermined value or the width W is less than or equal to a predetermined value and setting a flag (not shown) for the digital image data, Record that it is a target for subsequent processing.

  Here, it is assumed that the predetermined value is smaller than the predetermined value applied in step 110 of FIG. 6, and the predetermined width is larger than the predetermined width applied in step 114 of FIG. As a result, digital image data having a high possibility of including a person who could not be detected by the facility user measurement process is set as a subsequent processing target.

  In the next step 310, a measurement correction screen is displayed on the display 34 based on the digital image data flagged by the above processing.

  FIG. 14 shows an example of the measurement correction screen displayed on the display 34. As shown in the figure, in the measurement correction screen according to the present embodiment, for each digital image data flagged by the processing in step 306 in the left area of the screen, the image and event number in the determination area A1. Is displayed so as to be scrollable in the vertical direction of FIG.

  In the measurement correction screen, the entire image indicated by the digital image data designated by the operator of the facility user measurement device 20 is displayed in the right region of the screen, and the human image is displayed at the bottom of the display. An area for inputting designation of types (here, “unknown”, “patient”, “accompaniment”, and “employee”) and an entry to the examination room 62 are designated. An “IN” button, an “OUT” button designated when designating exit from the examination room 62, and an “end” button designated when designating the end of the measurement correction processing program are displayed. The

  When the measurement correction screen is displayed on the display 34, the operator refers to the displayed image of the determination area A1 and includes the person who enters the examination room 62 and the person who leaves the examination room 62 in the image. If it is, the mouse 32 designates the display area of the image or the event number corresponding to the image. In response to this, the entire image of the designated image is displayed in the right area of the display 34. FIG. 14 shows the display state when the image of event number 12 is designated.

  Next, the operator designates the display position indicating the type of person included in the designated image with the mouse 32, and when the person is going to enter the examination room 62, the operator presses the “IN” button. When the user wants to leave the examination room 62, the “OUT” button is designated by pointing with the mouse 32.

  The operator scrolls the image of the determination area A1 and the event number displayed in the left area of the display 34 by an operation with the arrow keys of the keyboard 30 or the mouse 32 as necessary. Further, the operator designates pointing with an “end” button with the mouse 32 when the measurement correction processing program is ended. The process for scrolling is executed by the CPU 22 as an interrupt process.

  Therefore, in the next step 312, the process waits until various designation inputs and operations as described above are performed by the operator, and in the next step 314, designation input or input / output of the type of person for the image designated by the operator is performed. It is determined whether or not the designated entry for leaving the room has been performed. If the determination is affirmative, the process proceeds to step 316, and the information designated by the operator is stored in the hard disk 28 by the facility user measurement process. After additional storage in the measurement information (see also FIG. 8), the process returns to step 312, and if a negative determination is made, the process proceeds to step 318.

  In step 318, it is determined whether or not the operator has designated the end of the main measurement correction processing program. If the determination is negative, the process returns to step 312. If the determination is positive, the main measurement correction processing program is determined. Exit.

  With this measurement correction processing program, it is possible to detect a considerable number of digital image data including people entering and leaving the room that could not be detected by the facility user measurement processing program.

  After the execution of the measurement correction process, the facility user measurement apparatus 20 according to the second embodiment executes an individual recognition process (see FIG. 9) similar to the first embodiment, and thereafter. Individual recognition correction processing is executed. Hereinafter, the individual recognition correction process will be described with reference to FIG. FIG. 15 is a flowchart showing the flow of the individual recognition correction processing program executed by the CPU 22 of the facility user measuring apparatus 20 when the individual recognition correction processing is executed. The program is also stored in a predetermined area of the ROM 26 in advance. It is remembered.

  First, in step 400, all weighted coincidences WC stored by the individual recognition process are read from the hard disk 28.

  In the next step 402 to step 406, digital image data having a matching degree WC equal to or higher than a predetermined value is detected, and a flag (not shown) for the digital image data is set to record that it is to be processed later.

  In the next step 408, an individual recognition correction screen is displayed on the display 34 based on the digital image data flagged by the above processing.

  FIG. 16 shows an example of the individual recognition correction screen displayed on the display 34. As shown in the figure, in the individual recognition correction screen according to the present embodiment, the attribute that enters the left area of the screen as the entry / exit attributes of the digital image data flagged by the processing in step 404. 16 is displayed so as to be scrollable in the vertical direction in FIG.

  Further, in the individual recognition correction screen, an image in which the attribute of leaving the room as the entry / exit attribute of the digital image data flagged by the processing of step 404 is added to the center area of the screen in FIG. Is scrollable.

  Further, in the individual recognition correction screen, the “association” button specified when the specified digital image data is associated with the right area of the screen and the end of the individual recognition correction processing program are specified. And an “END” button.

  When the individual recognition correction screen is displayed on the display 34, the operator includes an image group displayed in the left area (image when entering the room) and an image group displayed in the central area (image when leaving the room). , Two images including the same person are designated by pointing with the mouse 32. FIG. 16 shows the display state when the upper image is designated as the image when entering the room and the center image is designated as the image when leaving the room.

  Then, the operator designates the “association” button by pointing with the mouse 32 when associating each digital image data indicating the designated two images.

  Note that the operator scrolls the display image by an operation with an arrow key of the keyboard 30 or the mouse 32 as necessary. In addition, the operator designates pointing with the “end” button with the mouse 32 when the individual recognition correction processing program is ended. The process for scrolling is executed by the CPU 22 as an interrupt process.

  Therefore, in the next step 410, the operator waits for the designation of the “association” button or the “end” button, and in the next step 412, it is determined whether or not the “association” button has been designated. If YES in step 414, the flow advances to step 414, and the information for associating the digital image data indicating the two images designated by the operator with the association information stored in the hard disk 28 by the individual recognition process (also in FIG. 12). After the additional storage, the process returns to step 410, and if a negative determination is made, the process proceeds to step 416.

  In step 416, it is determined whether or not an “end” button has been designated by the operator. If a negative determination is made, the process returns to step 410. If an affirmative determination is made, the individual recognition correction processing program is ended. To do.

  The individual recognition correction processing program can detect a considerable number of digital image data including the same person that could not be detected by the individual recognition processing program.

  As described above in detail, in the present embodiment, in addition to being able to achieve the same effects as in the first embodiment, digital image data not extracted by the facility user measurement process and the individual recognition process In order to exit the examination room 62, the first designation that designates an image when a person is passing through the determination area A1 to enter the examination room 62 in the displayed image is displayed on the display 34. It is possible to input each designation of a second designation for designating an image when a person passes through the determination area A1 and a third designation for designating an image including the same person. Extraction omission due to processing and individual recognition processing can be compensated.

  In each of the above embodiments, the case where the “passing region” of the present invention is a single region (determination area A1) has been described, but the present invention is not limited to this, and for example, an examination room In the case where the entrance door and the exit door are individually provided in 62, it is also possible to adopt a form in which each passing area of a person with respect to these doors is applied as the “passing area” of the present invention. Further, when at least one of the entrance door and the exit door is provided in the examination room 62, each passing area of the person with respect to these doors is applied as the “passing area” of the present invention. You can also.

  In these cases, it is advantageous in terms of the cost for constructing the facility user measurement system that all the passing areas are photographed by one camera, but the present invention is not limited to this. When it is impossible to photograph with one camera, such as when they are relatively separated from each other or when they are on different wall surfaces, it is possible to adopt a form in which photographing is performed with a plurality of cameras. In these cases, each passing area is set in advance as a determination area similar to the present embodiment, and in the facility user measurement processing program (see FIG. 6), the determination area A1 is performed for each determination area. Process. Also in this case, the same effects as in the present embodiment can be obtained.

  In each of the above embodiments, the case where the area including the entrance / exit entrance 66 of the examination room 62 is applied as the determination area A1 has been described. However, the present invention is not limited to this. An area where the entrance / exit port 66 is not included in the vicinity of the exit port 66 and the patient passes when entering / exiting the examination room 62 may be used as the determination area A1. Also in this case, the same effects as in the present embodiment can be obtained.

  In each of the above embodiments, in the individual recognition process (see FIG. 9), whether or not they are the same person based on the degree of coincidence of the distribution states of the respective color components in the processing target entry image data and the processing target exit image data. However, the present invention is not limited to this. For example, the pattern of the person's clothes included in each image indicated by the processing target entry image data and the processing target exit image data matches. It is also possible to adopt a form in which the determination is made based on the degree or a form in which the determination is made based on the matching degree of the person's height. In addition, when applying the pattern matching degree of the clothes, the digital image data does not necessarily have to be acquired in color, and can be acquired in monochrome. In this case, the camera 50 can be made inexpensive, and the storage capacity required for storing digital image data can be reduced.

  Moreover, it is needless to say that the configuration of the facility user measurement system 10 (see FIGS. 1 to 3) shown in each of the above embodiments is an example, and can be appropriately changed without departing from the gist of the present invention. Yes.

  For example, in each of the above embodiments, a digital electronic still camera that captures a color still image for each preset period is applied as the camera 50. However, a digital camera that captures a color moving image is applied. You can also. Also in this case, the same effects as those of the above embodiments can be obtained.

  Furthermore, the flow of each process of the various processing programs (see FIGS. 6, 9, 13, and 15) shown in the above embodiments is also an example, and can be changed as appropriate without departing from the gist of the present invention. It goes without saying that it is possible.

It is a schematic diagram which shows the structure of the facility user measurement system 10 which concerns on embodiment. It is a block diagram which shows the structure of the electric system of the facility user measurement system 10 which concerns on embodiment. It is a see-through | perspective perspective view with which it uses for description of the installation state of the camera 50 which concerns on embodiment. It is the schematic which shows an example of the image 70 obtained by imaging | photography with the camera 50 which concerns on embodiment. It is the schematic where it uses for description of determination area A1 and brightness correction area A2 which concern on embodiment. It is a flowchart which shows the flow of a process of the facility user measurement process program which concerns on embodiment. It is the schematic where it uses for description of the detection method of the change direction of the image change area | region in the image shown by digital image data. It is a schematic diagram which shows an example of the measurement information which concerns on embodiment. It is a flowchart which shows the flow of a process of the separate recognition process program which concerns on embodiment. It is a schematic diagram which shows an example of the classification information which concerns on embodiment. It is a schematic diagram with which it uses for description of the extraction method of the feature data from digital image data. It is a schematic diagram which shows an example of the corresponding | compatible information which concerns on embodiment. It is a flowchart which shows the flow of a process of the measurement correction process program which concerns on embodiment. It is the schematic which shows an example of the measurement correction screen which concerns on embodiment. It is a flowchart which shows the flow of a process of the separate recognition correction process program which concerns on embodiment. It is the schematic which shows an example of the separate recognition correction screen which concerns on embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Facility user measurement system 20 Facility user measurement apparatus 22 CPU (1st extraction means, 2nd extraction means, determination means, exclusion means)
28 Hard disk (storage means)
32 mouse (designated input means)
34 Display (display means)
50 Camera (Image acquisition means)
A1 judgment area (passing area)
A2 Brightness correction area

Claims (13)

An image processing apparatus that performs image processing for grasping a person's entry / exit status with respect to a predetermined area,
Image acquisition means for capturing a plurality of pieces of image information by photographing a passing region through which a person passes when entering and exiting the predetermined region at predetermined intervals;
Comparing the plurality of image information acquired by the image acquisition means with background image information predetermined as image information acquired by the image acquisition means in a state where a person does not pass through the passage area; The image in which the moving direction in the time series order of the difference area on the image with respect to the background image information substantially matches the passing direction when a person passes through the passing area in order to enter the predetermined area. The information is extracted as first image information acquired by photographing when a person is passing through the passing area to enter the predetermined area, and the passing area is extracted so that the moving direction exits the predetermined area. the second image information person the passage region of the image information in order to exit from the predetermined area in the passing direction substantially coincides is acquired by shooting it passes through when a person passes A first extracting means for extracting,
In the direction from the newest first image information extracted by the first extracting means to the oldest first image information, the oldest first image information from the newest second image information extracted by the first extracting means. in order 2 image information indicating an image including the irradiation combined and the same person and the extracted second image information by the extracted first image information said first extraction means by said first extracting means Second extraction means for extracting the first image information and the second image information;
Storage means for storing acquisition times by the image acquisition means for each of the first image information and the second image information extracted by the second extraction means;
An image processing apparatus.   The first extraction unit calculates an absolute value of a difference between all the pixel data corresponding to the passing area of each of the plurality of pieces of image information acquired by the image acquisition unit and the corresponding pixel data of the background image information. Integration is performed for each of a plurality of pieces of image information, and the image information for which the integration result is a predetermined value or more is extracted.
  The image processing apparatus according to claim 1.   The first extraction unit compares each of the plurality of pieces of image information acquired by the image acquisition unit with the background image information, and the difference between the background image information and the difference area on the image is a predetermined width or less. Extract region
  The image processing apparatus according to claim 1.   The first extraction unit is configured to set the plurality of images so that the luminance level of the image indicated by the plurality of image information acquired by the image acquisition unit is a level corresponding to the luminance level of the image indicated by the background image information. Extracting the first image information and the second image information after adjusting the information
  The image processing apparatus according to any one of claims 1 to 3.   The second extraction unit divides a human image area included in each image indicated by the first image information and the second image information into an upper body area and a lower body area, and matches both the upper body area and the lower body area. The first image information and the second image information whose degrees are higher than a predetermined degree are extracted as indicating images containing the same person
  The image processing apparatus according to claim 1.   The image acquisition means acquires the image information in color,
  The second extraction means includes each color component of the upper body region of the person included in the image indicated by the first image information and each color component of the upper body region of the person included in the image indicated by the second image information. The degree of coincidence of the distribution state is higher than a predetermined degree, and the lower body region of the person included in the image indicated by the second image information and each color component of the lower body region of the person included in the image indicated by the first image information When the degree of coincidence of the distribution state with each color component is higher than a predetermined degree, the first image information and the second image information are extracted as indicating images containing the same person.
  The image processing apparatus according to claim 5.   The second extraction means includes a luminance region range that is equal to or greater than a predetermined value in a histogram of luminance values for each color component of the upper body region of the person included in the image indicated by the first image information, and the second image information. Indicated by the integrated value of the amount of overlap with a luminance area range that is equal to or greater than a predetermined value in a histogram of luminance values for each color component of the upper body area of the person included in the image indicated by Each color component of the lower body area of the person included in the image indicated by the luminance area range that is equal to or greater than a predetermined value in a histogram of luminance values for each color component of the lower body area of the person included in the image Both the integrated values of the overlapping amounts with the luminance region range that is equal to or greater than a predetermined value in the histogram of each luminance value are equal to or greater than the predetermined integrated value Extracting the first image information and the second image information as an indication of the image including the same person in some cases
  The image processing apparatus according to claim 6.   The second extraction means has a degree of coincidence between the pattern of the person's clothing included in the image indicated by the first image information and the pattern of the person's clothing included in the image indicated by the second image information based on a predetermined degree. The first image information and the second image information that are high are extracted as indicating images containing the same person.
  The image processing apparatus according to claim 1.   The second extraction unit is configured to store the first image information and the second image information so that luminance levels in the same image region of the image indicated by the first image information and the image indicated by the second image information are equal. Perform the extraction after adjusting at least one
  The image processing apparatus according to claim 1.   The storage means stores in advance exclusion feature amount information indicating a predetermined feature amount of an image of a person who wants to be excluded from the grasp target of the entry / exit status,
  The feature amount information indicating the feature amount of the human image indicated by the first image information and the second image information is collated with the excluded feature amount information, and the first image information and the second image information are used. Determining means for determining whether or not the person indicated is the person to be excluded;
  Exclusion means for excluding the first image information and the second image information determined to be excluded by the determination means from the processing target;
  The image processing apparatus according to claim 1, further comprising:   The storage means stores in advance clothing feature information indicating the features of the clothes of the person who wants to be excluded as the excluded feature amount information,
  The determination means collates the characteristics of the person's clothes indicated by the first image information and the second image information with the characteristics of the clothes indicated by the clothes characteristic information, and determines the first image information and the second image information. Determine whether the person indicated by the image information is the person who wants to exclude
  The image processing apparatus according to claim 10.   According to each image indicated by the first image information and the second image information not extracted by the first extraction means, and by the first image information and the second image information not extracted by the second extraction means. Display means for displaying each image shown;
  In the image displayed by the display means, a first designation that designates an image when a person is passing through the passing area to enter the predetermined area, and a person passing the passing area to exit the predetermined area. Designation input means for inputting at least one designation of a second designation for designating an image when
  Association instruction input means for inputting an association instruction for associating the image designated by the first designation and the image designated by the second designation, each including the same person;
  The first image information and the second image information indicating the image designated by the first designation and the image designated by the second designation associated by the association instruction input by the association instruction input means. The image processing apparatus according to claim 1, further comprising a control unit that performs control so as to store an acquisition time of each of the image acquisition units in the storage unit.   An image processing program for performing image processing for grasping a person's entry / exit status with respect to a predetermined area,
  An image acquisition step of capturing a plurality of pieces of image information by photographing a passing region through which a person passes when entering the predetermined region and exiting the predetermined region for each predetermined period;
  The plurality of image information acquired by the image acquisition step is compared with the background image information predetermined as the image information acquired by the image acquisition step in a state where no person passes through the passage area, The image in which the moving direction in the time series order of the difference area on the image with respect to the background image information substantially matches the passing direction when a person passes through the passing area in order to enter the predetermined area. The information is extracted as first image information acquired by photographing when a person is passing through the passing area to enter the predetermined area, and the passing area is extracted so that the moving direction exits the predetermined area. A second image acquired by photographing when a person is passing through the passing area in order to leave the image information that substantially matches the passing direction when the person passes through the predetermined area. A first extraction step of extracting as a broadcast,
  In the direction from the newest first image information extracted by the first extraction step to the oldest first image information, the oldest second image information extracted from the newest second image information extracted by the first extraction step. In the order of two image information, the first image information extracted by the first extraction step and the second image information extracted by the first extraction step are collated to indicate an image including the same person A second extraction step for extracting the first image information and the second image information;
  A storage step of storing, in a storage unit, an acquisition time of each of the first image information and the second image information extracted by the second extraction step by the image acquisition step;
  An image processing program for causing a computer to execute.

Images