JP2011081657A - Imaging system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain an imaging system which can image a detailed image of a mobile body while recording the movement of the mobile body by using one camera having an imaging optical system. <P>SOLUTION: A first CCD 212 has the number of pixels of a 1 megapixel class as well as a rectangular first imaging surface 217 on which an object image from an imaging lens 211 forms an image. The first CCD 212 is arranged inside a first terminal device 200 so that the direction of gravity for the first terminal device 200 matches the longitudinal direction of the first imaging surface 217. The first imaging surface 217 is divided into two by an upper part and a lower part of gravity. A portion arranged in the upper part of gravity forms an infrared ray imaging region and a portion arranged in the lower part of gravity forms a visible ray imaging region. The first CCD 212 images the object images that are imaged on the infrared ray imaging region and on the visible ray imaging region at once. The object at the upper part of a direction of gravity is made incident on the imaging lens 211 and is imaged in the visible ray imaging region and the object at the lower part of the direction of gravity is imaged in the infrared ray imaging region. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an imaging system that captures and records a moving object that meets a predetermined condition.

  Conventionally, a vehicle number reading system for reading a moving body, that is, a vehicle license plate, is known. The vehicle number reading system includes one camera and a position sensor attached to each traveling lane. When the position sensor detects the vehicle, an imaging command is issued to the camera provided on the traveling lane corresponding to the position sensor, and the camera that has received the imaging command captures a detailed image of the vehicle (Patent Document 1).

Japanese Patent Publication No. 2-38997

  However, in the conventional configuration, it is necessary to provide a camera in each traveling lane, and when a large number of traveling lanes must be monitored, the installation cost of the camera becomes great. Further, when it is necessary to record the movement of the vehicle, it is not possible to accurately capture the movement of the vehicle due to the limitation of the angle of view of the camera if there is one camera in one travel lane. On the other hand, in the configuration in which a plurality of cameras are provided in one travel lane, the installation cost is further increased because the cameras must be installed while adjusting the angle of view and the interlocking between the cameras.

  The present invention has been made in view of these problems, and provides an imaging system capable of capturing a detailed image of a moving object while recording the movement of the moving object using a single camera. With the goal.

  An imaging system according to the present invention has a first imaging area and a second imaging area that is different from the first imaging area, and moves in one direction across the first imaging area and the second imaging area. An image pickup unit for picking up a subject image, a color image pickup filter arranged in the first image pickup region, an invisible light image pickup filter arranged in the second image pickup region, and an image picked up by the image pickup unit are stored. The subject image matches a predetermined condition on the basis of either the first imaging area or the second imaging area, the storage means for recording, the recording means for recording the image stored in the storage means on the recording medium, and the first imaging area or the second imaging area. An imaging unit that captures the subject image that has passed through the color image filter and outputs a color image, and that captures the subject image that has passed through the invisible light image capturing filter. Output invisible light image When the determination unit determines that the subject image matches a predetermined condition, the recording unit records an image that has been captured and stored in the storage unit a predetermined period before the predetermined condition is met. It is characterized by that.

  The image processing apparatus further includes detection means for detecting whether or not the subject has moved from one of the first imaging area and the second imaging area to the other area, and the determination means matches the subject image with a predetermined condition. When the detection means detects that the subject has moved to the other area, it is preferable to cause the recording means to record an image that has been imaged and stored in the storage means a predetermined period before that time. .

  The imaging unit images a subject image that moves from the first imaging region to the second imaging region, and the determination unit determines whether the subject image meets a predetermined condition based on the color image, If the condition is met, an invisible light image of the subject image may be recorded by the recording means.

  Preferably, the determination unit causes the recording unit to record the invisible light image and the color image of the subject image when the predetermined condition is met.

  Preferably, the determination unit virtually divides the angle of view of the imaging unit into a plurality of regions, and determines that the predetermined condition is met when the subject image moves the region a plurality of times.

  The determination unit preferably determines that a predetermined condition is met when the subject image passes through a predetermined position at the angle of view of the imaging unit.

  The determination means may determine that a predetermined condition is met when the subject image has a predetermined color.

  Alternatively, the determination unit may determine that a predetermined condition is met when the subject image has a predetermined shape.

  The imaging means preferably has a first imaging area and a second imaging area provided on one imaging element.

  The image pickup means includes a first image pickup element and a second image pickup element, the first image pickup area is provided on the first image pickup element, and the second image pickup area is provided on the second image pickup element. May be.

  The imaging system is a vehicle monitoring system, the imaging means is fixed in the vicinity of the road and the angle of view is fixed, and the judging means meets a predetermined condition when the vehicle passes a predetermined range in the angle of view. You may judge.

  The imaging system is a vehicle monitoring system, the imaging means is fixed in the vicinity of the road and the angle of view is fixed, and the judging means virtually divides the angle of view into a plurality of areas, and the vehicle When moving, it may be determined that a predetermined condition is met.

  As described above, according to the present invention, an imaging system capable of capturing a detailed image of a moving body while recording the movement of the moving body using one camera is obtained.

It is a block diagram of the 1st imaging system by a 1st embodiment. It is a partial sectional view of the 1st image pick-up unit. It is the figure which showed the image which the 1st imaging unit imaged. It is a flowchart which shows a meandering recording process. It is a flowchart which shows a meandering initial set process. It is a flowchart which shows a roadside zone traffic recording process. It is a flowchart which shows a traffic initial set process. It is a flowchart which shows a vehicle color recording process. It is a flowchart which shows a vehicle color initial set process. It is a flowchart which shows a vehicle type recording process. It is a flowchart which shows a vehicle type initial set process. It is a block diagram of the 2nd imaging system by a 2nd embodiment. It is a partial sectional view of the 2nd image pick-up unit.

  Hereinafter, a first embodiment according to the present invention will be described with reference to FIGS.

  The first imaging system 100 includes a first terminal device 200 provided at a position where a subject image can be captured, and a central controller 300 provided at a position away from the first terminal device 200.

  The first terminal device 200 includes a first imaging device 210, a terminal calculation unit 220, a vehicle detection unit 230, a near infrared illumination unit 240, a terminal recording unit 250, and a terminal communication control unit 260. .

  The central controller 300 includes a central communication control unit 310, an alarm 320, a central recording unit 330, and a central processing unit 340.

  The first imaging device 210 includes an imaging lens 211 that constitutes an imaging optical system, a first CCD 212 that is an imaging element, a first visible light cut filter 213 that is an infrared imaging filter, and a first A polarizing filter 214, a first infrared cut filter 215 that is a color image capturing filter, and an optical path length adjustment filter 216 are provided.

  The imaging lens 211 has a focal length and a diaphragm for realizing a deep depth of field, and can perform pan-focus imaging. Accordingly, the imaging lens 211 can simultaneously focus on a subject existing at a distance from the first imaging device 210 of 10 m to 100 m.

  The first CCD 212 has a first imaging surface 217 having a rectangular shape on which a subject image from the imaging lens 211 is formed and has a number of pixels of 1 megapixel class or more. The first CCD 212 is provided inside the first terminal device 200 so that the direction of gravity of the first terminal device 200 coincides with the longitudinal direction of the first imaging surface 217. The first imaging surface 217 is divided into two equal parts above and below the gravity. A part provided above the gravity forms a near infrared imaging region 218, and a part provided below the gravity forms a visible light imaging region 219. The first CCD 212 captures the subject image formed on the first imaging surface 217, that is, the near-infrared imaging region 218 and the visible light imaging region 219 at a time. Imaging is performed, for example, 10 times per second, and the CCD 212 outputs a moving image consisting of 10 frames per second to the terminal calculation unit and the terminal recording unit 250. The imaging frame rate is appropriately set as necessary.

  The first polarizing filter 214 is an optical filter that transmits only light having a specific polarization state. The first visible light cut filter (near-infrared transmission filter) 213 is an optical filter that does not transmit visible light but transmits only near-infrared light included in infrared light, and forms an invisible light image capturing filter. The first polarizing filter 214 is attached to the first CCD 212 so as to cover the entire surface of the near infrared imaging region 218. A first visible light cut filter 213 is attached to the first CCD 212 so as to cover the entire surface of the first polarizing filter 214 and the near-infrared imaging region 218 and on the imaging lens 211 side of the first polarizing filter 214. It is done.

  The first infrared cut filter 215 is a filter that does not transmit infrared light but transmits light having other frequencies, and forms a color image capturing filter. The optical path length adjustment filter 216 is a filter that is provided to adjust the optical path length of transmitted light, and transmits light of all wavelengths. The optical path length adjustment filter 216 is attached to the first CCD 212 so as to cover the entire surface of the visible light imaging region 219. Then, on the imaging lens 211 side of the optical path length adjustment filter 216, the first infrared cut filter 215 is attached to the first CCD 212 so as to cover the entire surface of the optical path length adjustment filter and the visible light imaging region 219. The refractive index and thickness of the optical path length adjustment filter 216 are set such that the optical path length by the first infrared cut filter 215 and the optical path length adjustment filter 216 is the optical path length by the first visible light cut filter 213 and the first polarization filter 214. Is adjusted in consideration of the refractive index and thickness of each filter.

  For example, the first imaging device 210 is installed above the traveling lane for monitoring the traveling lane of an expressway, and photographs a vehicle approaching from a distance. The top and bottom of the subject image formed on the first imaging surface 217 is inverted from the top and bottom of the subject by the function of the imaging lens 211. For this reason, when a subject, for example, a car that travels and approaches, is located relatively far from the first imaging device 210, the subject image of the car is located in the visible light imaging region 219 below. . When the vehicle approaches and is positioned at a relatively short distance from the first imaging device 210, the subject image of the vehicle is positioned in the upper near infrared imaging region 218. Thereby, a color image is obtained for a subject located relatively far away, and an infrared image is obtained for a subject located relatively short distance (see FIG. 3).

  The vehicle detection unit 230 is a loop coil that is attached to a location within the imaging range of the first imaging device 210 and capable of detecting the position of the subject image formed in the near-infrared imaging region 218, and passes through the vehicle. Is detected. The terminal calculation unit 220 and the near-infrared illumination unit 240 are electrically connected, and a vehicle detection signal is transmitted to the terminal calculation unit 220 and the near-infrared illumination unit 240 when a vehicle is detected.

  The near-infrared illumination unit 240 is provided in a place where a vehicle traveling on a road can be illuminated and a position where a subject image formed in the near-infrared imaging region 218 can be illuminated, for example, a free-flow gate on a roadside band or road It is done. When a vehicle detection signal is received from the vehicle detection unit 230, near infrared light is emitted toward the vehicle.

  The terminal calculation unit 220 receives a vehicle detection signal from the vehicle detection unit 230 and a moving image from the first CCD 212 and performs image processing. In the image processing, meandering recording processing, roadside band traffic recording processing, vehicle color recording processing, and vehicle type recording processing described later are executed. In these recording processes, the terminal calculation unit 220 images the automobile registration number mark or vehicle number mark (so-called license plate) imaged in the near-infrared imaging region 218 and the driver to create a still image, and the terminal recording unit 250.

  The terminal recording unit 250 is a storage unit including a semiconductor storage device such as a hard disk or a DRAM, and temporarily stores a moving image received from the first imaging device 210 and a still image received from the terminal calculation unit 220.

  The terminal communication control unit 260 is electrically connected to the central controller 300 by wired or wireless communication. Then, among the still images and moving images recorded by the terminal recording unit 250, those that meet the conditions described later in each recording process are transmitted to the central controller 300.

  The central communication control unit 310 included in the central controller 300 receives still images and moving images and transmits them to the central recording unit 330. When a still image or a moving image is received from the terminal communication control unit 260, an alarm signal is transmitted to the alarm 320.

  The central recording unit 330 is a recording unit such as a hard disk, and records still images and moving images received from the central communication control unit 310.

  The alarm 320 is a speaker, a display, or a warning light. When an alarm signal is received from the central communication control unit 310, the alarm 320 emits a warning sound, displays a warning image, or blinks a warning light.

  The central processing unit 340 reads the vehicle registration number of the vehicle from the license plate included in the still image, and determines whether or not it matches the monitoring target registration number.

  Next, the meandering recording process will be described with reference to FIGS. The meander recording process is started by the terminal calculation unit 220 when the first terminal device 200 can start photographing.

  In step S401, a meandering initial setting process is executed. In the meandering initial set process, an area occupied by the travel lane, that is, a travel lane area is extracted from the image captured by the first imaging device 210, a lane number is assigned to the extracted travel lane area, and one frame is extracted from the moving image. The details will be described later. The traveling lane refers to an area on which a vehicle passes on a road, which is divided by a white line, a yellow line, or the like with a predetermined width. The operation of extracting one frame from the moving image is called capture, and the frame captured in the meandering initial set process is called the 0th frame.

  In step S402, initialization is performed by assigning 0 to the lane change count parameter m. The lane change frequency parameter m is a parameter indicating the number of lane changes made by the vehicle that is the observation target.

  In step S403, initialization is performed by substituting 1 into the frame number parameter n. The frame number parameter n is a number that is continuously assigned to frames included in the moving image output by the first imaging device 210.

  In step S404, the (n + 1) th and n + 2th frames are captured from the moving image. The (n + 2) th frame is a frame acquired after a predetermined time from the (n + 1) th frame. The predetermined time is, for example, 0.1 seconds.

  In step S405, the difference P between the 0th frame or the nth frame and the (n + 1) th frame and the difference Q between the (n + 2) th frame and the (n + 1) th frame are calculated. Thereby, pixels having different values are extracted between a specific frame and a frame after a predetermined time from the frame. Pixels that differ between frames represent a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object.

  In step S406, the difference P and the difference Q are binarized using a predetermined threshold. Then, the logical product R of the binarized difference P and the binarized difference Q is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product R of the binarized difference P and difference Q, it is possible to specify a moving body included in a specific frame and a frame after a predetermined time. This specific operation is called marking.

  In step S407, the travel lane area where the moving body marked in step S406 is located and the lane number of the travel lane area are calculated. Thereby, the traveling lane in which the mobile body is currently located is specified.

  In step S408, one frame is captured from the moving image output from the first imaging device 210. Each time a frame is captured, the frame number parameter n is incremented by one. Thus, three frames are captured together with the two frames captured in step S404. In the following description, it is assumed that the most recently captured frame is the (n + 2) th frame, the next most recently captured frame is the (n + 1) th frame, and the next most recently captured frame is the nth frame.

  In step S409, the difference S between the nth frame and the (n + 1) th frame and the difference T between the (n + 2) th frame and the (n + 1) th frame are calculated. As a result, a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object is extracted.

  In step S410, the difference S and the difference T are binarized using a predetermined threshold. Then, the logical product U of the binarized difference S and the binarized difference T is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product U of the binarized difference S and difference T, it is possible to mark a moving body included in a specific frame and a frame after a predetermined time.

  In step S411, the travel lane area where the moving body marked in step S410 is located and the lane number of the travel lane area are calculated. Thereby, the traveling lane in which the moving body is currently located is specified.

  In step S412, it is determined whether the lane number calculated in step S407 is different from the lane number calculated in step S411. If they are different, the process proceeds to step S413. If they are the same, the process proceeds to step S414.

  In step S413, the lane change count parameter m is increased by one. This is because if the lane number is different in step S412, it is considered that the moving body has changed the traveling lane.

  In step S414, it is determined whether or not the lane change number parameter m is equal to or greater than the maximum lane change number mMAX. The maximum number of lane changes mMAX is the maximum value of the number of lane changes allowed for the moving body, and is determined by the user before executing the meander recording process. If the lane change number parameter m is greater than or equal to the maximum lane change number mMAX, the process proceeds to step S415; otherwise, the process proceeds to step S408.

  In step S415, a plurality of frames are captured from the moving image received from the first imaging device 210. The imaging intervals and the number of frames are determined so as to record all travel lane changes made by the moving body. In the present embodiment, 10 frames are captured every 0.1 second.

  In step S416, all the frames captured in step S415 are transmitted to the central controller 300 via the terminal communication control unit 260. In the central controller 300, all received frames are stored in the central recording unit 330. Then, the process ends.

  The meandering initial setting process will be described.

  In step S501, the area occupied by the travel lane, that is, the travel lane area is extracted from the image captured by the first imaging device 210. The travel lane area is recognized when the terminal calculation unit 220 performs image processing on a white line or a yellow line that separates the travel lane and the travel lane. Further, in order to detect a vehicle passing through the roadside belt, the roadside belt is also recognized as one traveling lane area.

  In the next step S502, a number is assigned to the extracted travel lane area. When the image shown in FIG. 3 is obtained, the roadside zone located on the right side of the traveling lane in which the vehicle imaged in the near-infrared image is present is the lane number L1, and the rightmost image in which the vehicle imaged in the near-infrared image is present. Is the lane number L2, the left lane is the lane number L3, and the leftmost lane where there is no imaged vehicle is the lane number L4.

  In the next step S503, one frame is extracted from the moving image. By capturing an image of a traveling lane in which an observation target does not exist in advance, it is easy to detect a moving object. Then, the process ends.

  According to the meander recording process, it is possible to automatically image and record a vehicle that repeats the lane change for the maximum number of lane changes mMAX or more.

  Next, the roadside band traffic recording process will be described with reference to FIGS. The roadside band traffic recording process is started by the terminal calculation unit 220 when the first terminal device 200 can start shooting.

  In step S601, a pass initial setting process is executed. The traffic initial setting process is a process of assigning a lane number to the traveling lane area, setting the maximum roadside band traffic count mMAX, extracting one frame from the moving image, and recognizing the roadside band, which will be described in detail later. . The maximum roadside band traffic number mMAX is the number of times the mobile body has passed the roadside band, more specifically the maximum value of the period, and is determined by the user. Note that the frame captured in the passing initial set process is referred to as the 0th frame.

  In step S602, initialization is performed by substituting 0 for the confirmation count parameter m. The confirmation count parameter m is a parameter indicating the number of times that the vehicle to be observed has passed the roadside belt, more specifically, the period.

  In step S603, initialization is performed by assigning 1 to the frame number parameter n. The frame number parameter n is a number that is continuously assigned to frames included in the moving image output by the first imaging device 210.

  In step S604, the (n + 1) th and n + 2th frames are captured from the moving image. The (n + 2) th frame is a frame acquired after a predetermined time from the (n + 1) th frame. The predetermined time is, for example, 0.1 seconds.

  In step S605, the difference P between the 0th frame or the nth frame and the (n + 1) th frame and the difference Q between the (n + 2) th frame and the (n + 1) th frame are calculated. Thereby, pixels having different values are extracted between a specific frame and a frame after a predetermined time from the frame. Pixels that differ between frames represent a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object.

  In step S606, the difference P and the difference Q are binarized using a predetermined threshold. Then, the logical product R of the binarized difference P and the binarized difference Q is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product R of the binarized difference P and difference Q, it is possible to specify a moving body included in a specific frame and a frame after a predetermined time.

  In step S607, the frame number parameter n is incremented by one.

  In step S608, one frame is captured from the moving image output by the first imaging device 210. As a result, three frames are captured together with the two frames captured in step S604. In the following description, it is assumed that the most recently captured frame is the (n + 2) th frame, the next most recently captured frame is the (n + 1) th frame, and the next most recently captured frame is the nth frame.

  In step S609, the difference S between the nth frame and the (n + 1) th frame and the difference T between the (n + 2) th frame and the (n + 1) th frame are calculated. As a result, a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object is extracted.

  In step S610, the difference S and the difference T are binarized using a predetermined threshold. Then, the logical product U of the binarized difference S and the binarized difference T is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product U of the binarized difference S and difference T, it is possible to mark a moving body included in a specific frame and a frame after a predetermined time.

  In step S611, the frame number parameter n is incremented by one.

  In step S612, it is determined whether or not the mobile object is located in the roadside band. The determination is made by performing image processing to determine whether or not the moving body marked in step S610 is present in the travel lane area corresponding to the roadside zone. If the moving body is located in the roadside belt, the process proceeds to step S613; otherwise, the process returns to step S608.

  In step S613, the confirmation count parameter m is incremented by one. This is because the moving body is located in the roadside band in step S612.

  In step S614, it is determined whether or not the confirmation count parameter m is equal to or greater than the maximum roadside band traffic confirmation count mMAX. Since the frames are acquired at a predetermined cycle, the time during which the moving body exists in the roadside band can be measured by counting the number of frames obtained by imaging the moving body located in the roadside band. If the confirmation count parameter m is equal to or greater than the maximum roadside band traffic confirmation count mMAX, the process proceeds to step S615; otherwise, the process returns to step S608.

  In step S615, it is determined using the signal from the vehicle detection unit 230 whether the moving body is located at the shooting point. When the loop coil that forms the vehicle detection unit 230 is installed at the photographing point and a vehicle detection signal transmitted by the loop coil is received, it is determined that the moving body is located at the photographing point. If the moving body is located at the shooting point, the process proceeds to step S616; otherwise, the process returns to step S608.

  In step S616, a plurality of frames are captured from the moving image received from the first imaging device 210. The imaging intervals and the number of frames are determined so as to record all lane changes made by the moving body. In the present embodiment, 10 frames are captured every 0.1 second.

  In step S617, all the frames captured in step S616 are transmitted to the central controller 300 by the terminal communication control unit 260. Then, the process ends.

  The traffic initial set process will be described.

  In step S701, the area occupied by the travel lane, that is, the travel lane area is extracted from the image captured by the first imaging device 210. The travel lane area is recognized when the terminal calculation unit 220 performs image processing on a white line or a yellow line that separates the travel lane and the travel lane. Further, in order to detect a vehicle passing through the roadside belt, the roadside belt is also recognized as one traveling lane area. Then, a number is assigned to the extracted travel lane area. When the image shown in FIG. 3 is obtained, the roadside zone located on the right side of the traveling lane in which the vehicle imaged in the near-infrared image is present is the lane number L1, and the rightmost image in which the vehicle imaged in the near-infrared image is present. Is the lane number L2, the left lane is the lane number L3, and the leftmost lane where there is no imaged vehicle is the lane number L4.

  In step S702, the maximum roadside band traffic confirmation count mMAX is acquired.

  In the next step S703, one frame is extracted from the moving image. By capturing an image of a traveling lane in which an observation target does not exist in advance, it is easy to detect a moving object.

  In step S704, the area constituting the roadside belt is recognized as the roadside belt area. Then, the process ends.

  According to the roadside band traffic recording process, it is possible to automatically capture and record a vehicle existing in the roadside band for the maximum roadside band traffic confirmation count mMAX or more.

  Next, the vehicle color recording process will be described with reference to FIGS. The vehicle color recording process is started by the terminal calculation unit 220 when the first terminal device 200 can start shooting.

  In step S801, a vehicle color initial set process is executed. The vehicle color initial set process is a process for setting a vehicle color that is symmetrical to the detection, and will be described in detail later. The vehicle color that becomes detection symmetry is determined by the user.

  In step S802, initialization is performed by assigning 0 to the frame number parameter n. The frame number parameter n is a number that is continuously assigned to frames included in the moving image output by the first imaging device 210.

  In step S803, the (n + 1) th and n + 2th frames are captured from the moving image. The (n + 2) th frame is a frame acquired after a predetermined time from the (n + 1) th frame. The predetermined time is, for example, 0.1 seconds.

  In step S804, the difference P between the nth frame and the (n + 1) th frame and the difference Q between the (n + 2) th frame and the (n + 1) th frame are calculated. Thereby, pixels having different values are extracted between a specific frame and a frame after a predetermined time from the frame. Pixels that differ between frames represent a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object.

  In step S805, the difference P and the difference Q are binarized using a predetermined threshold. Then, the logical product R of the binarized difference P and the binarized difference Q is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product R of the binarized difference P and difference Q, it is possible to specify a moving body included in a specific frame and a frame after a predetermined time.

  In step S806, the n + 1-th frame is added to the logical product R calculated for the marked moving body to calculate a color recognition image C having color information.

  In step S807, it is determined whether or not the vehicle color of the moving object matches the vehicle color (target color) to be searched. If the vehicle color of the moving object matches the color to be searched, the process proceeds to step S808; otherwise, the process returns to step S802.

  In step S808, one frame is captured from the moving image output by the first imaging device 210. The frame captured here is a color image. Thereby, a color image is recorded in the terminal recording part 250, and the color of the vehicle body of a moving body can be recognized.

  In step S809, the frame number parameter n is incremented by one.

  In step S810, one frame is captured from the moving image output by the first imaging device 210. As a result, four frames are captured together with the three frames captured in steps S803 and S808. Of these, the following processing is performed using the three most recently captured frames. In the following description, it is assumed that the most recently captured frame is the (n + 2) th frame, the next most recently captured frame is the (n + 1) th frame, and the next most recently captured frame is the nth frame.

  In step S811, the difference S between the nth frame and the (n + 1) th frame and the difference T between the (n + 2) th frame and the (n + 1) th frame are calculated. As a result, a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object is extracted.

  In step S812, the difference S and the difference T are binarized using a predetermined threshold. Then, the logical product U of the binarized difference S and the binarized difference T is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product U of the binarized difference S and difference T, it is possible to mark a moving body included in a specific frame and a frame after a predetermined time.

  In step S813, it is determined using the signal from the vehicle detection unit 230 whether the moving body is located at the shooting point. When the loop coil that forms the vehicle detection unit 230 is installed at the photographing point and a vehicle detection signal transmitted by the loop coil is received, it is determined that the moving body is located at the photographing point. If the moving body is located at the shooting point, the process proceeds to step S814; otherwise, the process returns to step S809.

  In step S814, a plurality of frames are captured from the moving image received from the first imaging device 210. The imaging intervals and the number of frames are determined so as to record all lane changes made by the moving body. In the present embodiment, 10 frames are captured every 0.1 second.

  In step S815, all the frames captured in step S814 are transmitted to the central controller 300 by the terminal communication control unit 260. Then, the process ends.

  The car color initial set process will be described.

  In step S901, the user sets a vehicle color that is symmetrical to the detection. The vehicle color is set using HSV (hue, saturation, brightness), RGB (red, green, blue) color space, or the like. After the user sets, the process ends.

  According to the vehicle color recording process, a vehicle having a desired vehicle color can be automatically imaged and recorded.

  Next, the vehicle type recording process will be described with reference to FIGS. The vehicle type recording process is started by the terminal calculation unit 220 when the first terminal device 200 can start shooting.

  In step S1001, a vehicle type initial set process is executed. The vehicle type initial set process is a process of setting the vehicle type information, setting the maximum vehicle type detection count mMAX, and extracting one frame from the moving image, and will be described in detail later. The maximum vehicle type detection count mMAX is the number of frames in which a moving body corresponding to a desired vehicle type (target vehicle type) is imaged, and is determined by the user. Note that the frame captured in the passing initial set process is referred to as the 0th frame.

  In step S1002, initialization is performed by substituting 0 for the confirmation count parameter m. The confirmation number parameter m is a parameter indicating the number of times a moving body corresponding to a desired vehicle type is imaged.

  In step S1003, initialization is performed by substituting 1 into the frame number parameter n. The frame number parameter n is a number that is continuously assigned to frames included in the moving image output by the first imaging device 210.

  In step S1004, the (n + 1) th and n + 2th frames are captured from the moving image. The (n + 2) th frame is a frame acquired after a predetermined time from the (n + 1) th frame. The predetermined time is, for example, 0.1 seconds.

  In step S1005, the difference P between the nth frame and the (n + 1) th frame and the difference Q between the (n + 2) th frame and the (n + 1) th frame are calculated. Thereby, pixels having different values are extracted between a specific frame and a frame after a predetermined time from the frame. Pixels that differ between frames represent a subject image that has moved from the moment of capturing a specific frame to the moment of capturing a frame after a predetermined time, that is, a moving object.

  In step S1006, the difference P and the difference Q are binarized using a predetermined threshold. Then, the logical product R of the binarized difference P and the binarized difference Q is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product R of the binarized difference P and difference Q, it is possible to specify a moving body included in a specific frame and a frame after a predetermined time.

  In step S1007, the frame number parameter n is incremented by one.

  In step S1008, one frame is captured from the moving image output by the first imaging device 210. As a result, three frames are captured together with the two frames captured in step S1004. In the following description, it is assumed that the most recently captured frame is the (n + 2) th frame, the next most recently captured frame is the (n + 1) th frame, and the next most recently captured frame is the nth frame.

  In step S1009, the frame number parameter n is incremented by one.

  In step S1010, a difference S between the nth frame and the (n + 1) th frame and a difference T between the (n + 2) th frame and the (n + 1) th frame are calculated. As a result, a subject image that has moved from the moment when a specific frame is imaged to the moment when a frame after a predetermined time is imaged, that is, a moving object is extracted.

  In step S1011, the difference S and the difference T are binarized using a predetermined threshold. Then, the logical product U of the binarized difference S and the binarized difference T is calculated. By binarizing using a predetermined threshold, pixels that do not represent a moving object, that is, noise can be removed. By taking the logical product U of the binarized difference S and difference T, it is possible to mark a moving body included in a specific frame and a frame after a predetermined time.

  In step S1012, it is determined whether or not the mobile object corresponds to a desired vehicle type. The determination is made by performing image processing on whether or not the mobile object marked in step S1010 corresponds to the target vehicle type. If the mobile object corresponds to the target vehicle type, the process proceeds to step S1013; otherwise, the process proceeds to step S1014.

  In step S1013, the confirmation count parameter m is incremented by one. This is because the mobile body corresponds to the target vehicle type in step S1012.

  In step S1014, it is determined whether or not the confirmation number parameter m is equal to or greater than the maximum vehicle type detection number mMAX. The number of frames in which the moving object corresponding to the target vehicle type is imaged is counted, and it is determined that the moving object corresponds to the target vehicle type when the number is equal to or greater than the maximum vehicle type detection count mMAX. Thereby, the accuracy of target vehicle type detection can be raised. If the confirmation count parameter m is equal to or greater than the maximum vehicle type detection count mMAX, the process proceeds to step S1015; otherwise, the process returns to step S1008.

  In step S1015, it is determined using the signal from the vehicle detection unit 230 whether the moving body is located at the shooting point. When the loop coil that forms the vehicle detection unit 230 is installed at the photographing point and a vehicle detection signal transmitted by the loop coil is received, it is determined that the moving body is located at the photographing point. If the moving body is located at the shooting point, the process proceeds to step S1016; otherwise, the process returns to step S1008.

  In step S1016, a plurality of frames are captured from the moving image received from the first imaging device 210. The imaging intervals and the number of frames are determined so as to record all lane changes made by the moving body. In the present embodiment, 10 frames are captured every 0.1 second.

  In step S1017, the terminal communication control unit 260 transmits all frames captured in step S1016 to the central controller 300. Then, the process ends.

  The vehicle type initial set process will be described.

  In step S1101, the user inputs a vehicle type that is symmetrical to the detection.

  In step S1102, the maximum vehicle type detection count mMAX is acquired.

  In the next step S1103, one frame is extracted from the moving image. By capturing an image of a traveling lane in which an observation target does not exist in advance, it is easy to detect a moving object. Then, the process ends.

  According to the vehicle type recording process, a vehicle corresponding to a desired vehicle type can be automatically imaged and recorded.

  According to the present embodiment, it is possible to capture a detailed image of the moving body while monitoring the movement of the moving body.

  In the present embodiment, as shown in FIG. 2, the near-infrared imaging region 218 and the visible light imaging region 219 are divided into two equal parts at the position where the optical axis of the imaging lens 211 passes. Accordingly, the ratio of one imaging region may be increased.

  In addition, the received signal of the loop coil is used as a process for determining whether or not the moving body is located at the shooting point. Instead, the moving speed and moving direction of the extracted moving body image on the image sensor surface are used. May be calculated, and near-infrared imaging may be performed by predicting the timing of reaching the area corresponding to the imaging point in the near-infrared imaging area 218 based on the moving speed and the moving direction.

  Next, a second embodiment according to the present invention will be described with reference to FIGS. The same components as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  The second imaging system 400 includes a second terminal device 500 provided at a position where a subject image can be captured, and a central controller 300 provided at a position away from the second terminal device 500.

  The second terminal device 500 includes a second imaging device 510, a terminal calculation unit 220, a vehicle detection unit 230, a near infrared illumination unit 240, a terminal recording unit 250, and a terminal communication control unit 260. .

  The second imaging device 510 includes an imaging lens 211, a second CCD 512 and a third CCD 515 that are imaging elements, a second visible light cut filter 516 that is an infrared imaging filter, and a second polarization. A filter 517, a second infrared cut filter 514 that is a color image capturing filter, and a half mirror 513 are provided.

  The second and third CCDs 512 and 515 have a number of pixels of 1 megapixel class or more and have rectangular second and third imaging surfaces 519 and 520 on which the subject image from the imaging lens 211 is formed. .

  The second lens unit 500 is arranged so that the optical axis direction of the imaging lens 211 is parallel to the short side direction of the second imaging surface 519 and the gravity direction of the second terminal device 500 is perpendicular to the second imaging surface 519. A second CCD 512 is provided inside the second terminal device 500. Further, the optical axis direction of the imaging lens 211 is perpendicular to the third imaging surface 520, and the gravity direction of the second terminal device 500 is coincident with the short side direction of the third imaging surface 520. A third CCD 515 is provided inside the second terminal device 500. The second imaging surface 519 forms a visible light imaging region, and the third imaging surface 520 forms a near infrared imaging region. The second and third CCDs 512 and 515 simultaneously capture the subject images formed on the second and third imaging surfaces 519 and 520, that is, the visible light imaging region 519 and the near infrared imaging region 520. The second and third CCDs 512 and 515 output a moving image of 10 frames per second to the terminal calculation unit 220 and the terminal recording unit 250.

  The second polarizing filter 517 is an optical filter that transmits only light having a specific polarization state, and is attached to the third CCD 515 so as to cover the entire surface of the near-infrared imaging region 520.

  The second visible light cut filter 516 is an optical filter that transmits only near infrared rays included in infrared rays. A second visible light cut filter 516 is attached to the imaging lens 211 side of the second polarizing filter 517 so as to cover the entire surface of the second polarizing filter 517.

  The second infrared cut filter 514 is a filter that does not transmit infrared light but transmits light having other frequencies, and is attached to the second CCD 512 so as to cover the entire surface of the visible light imaging region 519. The refractive index or thickness of the second infrared cut filter 514 is such that the optical path length by the second infrared cut filter 514 is equal to the optical path length by the second visible light cut filter 516 and the second polarizing filter 517. To be adjusted.

  The half mirror 513 is provided on the optical path and between the imaging lens 211 and the second CCD 512 and the third CCD 515 so as to be 45 degrees with respect to the optical axis of the imaging lens 211. The half mirror 513 reflects half of the incident light and transmits the remaining half. Light from a distant view is incident on the imaging lens 211 and then reflected by the half mirror 513 to form an image on the visible light image capturing area 519, and light from the near view is transmitted through the half mirror 513 to be connected to the near infrared image capturing area 520. Image. The visible light imaging region 519 and the near-infrared imaging region 520 are perpendicular to the optical path.

  With these configurations, the second CCD 512 captures a distant subject with a color image, and the third CCD 515 captures a near subject with a near-infrared image (see FIG. 3).

  According to the present embodiment, it is possible to use an inexpensive and small image sensor. In addition, as in the first embodiment, it is possible to save the trouble of providing a separate filter by dividing the imaging surface of one imaging device into a plurality of parts.

  In any of the embodiments, the imaging element is not limited to the CCD, and may be a CMOS or the like. Further, the number of pixels of the CCD is not limited to 1 megapixel class or more. Furthermore, imaging is not limited to 10 times per second.

  By adjusting the refractive index or thickness of the infrared cut filter without using the optical path length adjustment filter 216, the optical path length of the visible light cut filter and the polarization filter is made equal to the optical path length of the infrared cut filter. Also good.

  Further, the vehicle detection unit 230 may not be a loop coil, and may be any device that can detect the position of the vehicle, such as an infrared sensor or a sound wave sensor.

  In addition, it can also make a fruit into a monitoring object instead of a vehicle. The fruit can be selected by using the sugar content of the fruit instead of the color as the discrimination condition. Further, by using the type and size of the fruit instead of the vehicle type, it is possible to select different kinds of fruits.

  Further, the parcel may be monitored instead of the vehicle. Parcels can be sorted according to destination, size, and shape.

DESCRIPTION OF SYMBOLS 100 1st imaging system 200 Terminal device 210 1st imaging device 211 Imaging lens (imaging optical system)
212 First CCD
213 First visible light cut filter 214 First polarizing filter 215 First infrared cut filter 216 Optical path length adjustment filter 217 First imaging surface 218 Near infrared imaging region 219 Visible light imaging region 220 Terminal calculation unit 230 Vehicle detection Unit 240 Near-infrared illumination unit 250 Terminal recording unit 260 Terminal communication control unit 300 Central controller 310 Central communication control unit 320 Alarm 330 Central recording unit 340 Central processing unit

Claims (12)

An image of a subject having a first imaging area and a second imaging area different from the first imaging area and moving in one direction across the first imaging area and the second imaging area Imaging means to perform,
A color image capturing filter disposed in the first imaging region; an invisible light image capturing filter disposed in the second imaging region;
Storage means for storing an image captured by the imaging means;
Recording means for recording an image stored in the storage means on a recording medium;
Determining means for determining whether or not the subject image meets a predetermined condition based on one of the images of the first imaging region or the second imaging region;
The imaging means captures a subject image that has passed through the color image filter and outputs a color image, captures a subject image that has passed through the invisible light image capture filter, and outputs an invisible light image;
If it is determined that the subject image matches a predetermined condition, the determination unit captures an image captured and stored in the storage unit a predetermined period before the predetermined condition is met. An imaging system characterized by having a means for recording. A detection unit that detects whether or not the subject has moved from one of the first imaging region and the second imaging region to the other region;
The determination means is an image picked up before a predetermined period from when the subject image meets a predetermined condition and when the detection means detects that the subject has moved to the other area. The imaging system according to claim 1, wherein the recording unit records an image stored in the recording unit. The imaging means captures a subject image moving from the first imaging area to the second imaging area,
The determination unit determines whether or not the subject image matches a predetermined condition based on the color image, and if the predetermined image satisfies the predetermined condition, causes the recording unit to record an invisible light image of the subject image. The imaging system according to claim 1 or 2, wherein   4. The imaging system according to claim 3, wherein the determination unit causes the recording unit to record an invisible light image and a color image of the subject image when the predetermined condition is satisfied.   The determination unit virtually divides the angle of view of the imaging unit into a plurality of regions, and determines that a subject image meets a predetermined condition when the subject image moves the region a plurality of times. 5. The imaging system according to 4.   The imaging system according to claim 1, wherein the determination unit determines that a predetermined condition is met when a subject image passes through a predetermined position at an angle of view of the imaging unit.   The imaging system according to claim 1, wherein the determination unit determines that a predetermined condition is met when the subject image has a predetermined color.   The imaging system according to claim 1, wherein the determination unit determines that a predetermined condition is met when the subject image has a predetermined shape.   The imaging system according to claim 1, wherein the imaging unit includes the first imaging area and the second imaging area provided on one imaging element.   The imaging means includes a first imaging element and a second imaging element, the first imaging area is provided on the first imaging element, and the second imaging area is the second imaging area. The imaging system according to claim 1, wherein the imaging system is provided on an imaging element. The imaging system is a vehicle monitoring system,
The imaging means is fixed in the vicinity of the road and the angle of view is fixed,
The imaging system according to claim 1, wherein the determination unit determines that a predetermined condition is met when the vehicle passes a predetermined range in the angle of view. The imaging system is a vehicle monitoring system,
The imaging means is fixed in the vicinity of the road and the angle of view is fixed,
The imaging system according to claim 1, wherein the determination unit virtually divides the angle of view into a plurality of regions and determines that a predetermined condition is met when the vehicle moves the region a plurality of times.

Images