JP2005025328A - Congestion monitoring system and congestion monitoring program - Google Patents

Abstract

An object of the present invention is to enable measurement of the number of people present in a predetermined area.
Image data corresponding to an image of a predetermined area photographed by a camera is accumulated as a still image DB via a control unit. The control unit 200 reads a background image corresponding to the region from the background image DB 30, compares the pixel value for each corresponding pixel to obtain a difference in pixel value, and obtains an area where a human exists. Then, weighting according to the distance from the camera 100 is performed on the area based on the weighting table 50, and the number of persons in the region is determined based on the area (difference from the background image) and the number of persons estimation DB40. presume.
[Selection] Figure 1

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a congestion monitoring system and a congestion monitoring program, and more particularly to a congestion monitoring system and a congestion monitoring program for monitoring the degree of congestion by measuring the number of people in a predetermined area such as a public facility, road, or park. .
[0002]
[Prior art]
For example, when an image is taken from the upper side of an elevator to grasp the number of people, as the number of people in the elevator increases, the shade becomes darker and erroneously recognized as a person. , See Patent Document 1).
In addition, there is one that verifies an image using computer graphics (for example, see Patent Document 2).
[0003]
[Patent Document 1]
Japanese Patent Laid-Open No. 11-259625 [Patent Document 2]
Japanese Patent Laid-Open No. 2000-34691
[Problems to be solved by the invention]
However, in the inventions described in Patent Document 1 and Patent Document 2, when using a camera for photographing outdoors, it is desirable to photograph the range of the photographing object from directly above, but the installation location is often limited. Therefore, it is difficult to install the camera at a position where it can be photographed from directly above the position of the center of gravity of the subject of photographing, and there are cases where photographing is performed from an oblique direction with respect to the position of the center of gravity. In this case, it is difficult to grasp a person based on an image having a depth. That is, as the shooting distance from the camera increases, the size of the photographed person differs, so that there is a case where the person is not counted and an error occurs. Further, when photographing from an oblique direction, there is a problem that if there is a person behind the person, the person in the back may be hidden behind the person in front and may not be photographed by the camera.
[0005]
The present invention has been made in view of such a situation, and a congestion monitoring system and a congestion monitoring program capable of estimating the number of people in a predetermined area with higher accuracy even when photographing from an oblique direction. The purpose is to provide.
[0006]
[Means for Solving the Problems]
In order to solve the above-described problem, the present invention provides an imaging unit that captures an image of a predetermined region, a first storage unit that stores the image captured by the imaging unit, and a reference that is captured in advance. The second storage means for storing the reference image of the region, the value of each pixel constituting the image and the value of each pixel constituting the reference image are compared with each other, and the pixel value A difference calculating means for calculating a difference, a counting means for counting the number of pixels in which a difference in pixel values calculated by the difference calculating means exceeds a predetermined reference value, and a pixel counted by the counting means Pixel number correcting means for weighting the number of pixels according to the position of each pixel, and estimating means for detecting the number of people in the region based on the number of pixels weighted by the pixel number correcting means. Preparation And wherein the Rukoto.
[0007]
In addition, the present invention provides an imaging unit that captures an image of a predetermined region, a first storage unit that stores the image captured by the imaging unit, and a reference image of the region that serves as a reference captured in advance. The second storage means for storing, the value of each pixel constituting the image, and the value of each pixel constituting the reference image are compared with each other corresponding to each other, and a difference calculating means for calculating a difference between the pixel values Counting means for counting the number of pixels in which the difference in pixel values calculated by the difference calculating means exceeds a predetermined reference value; and the number of pixels counted by the counting means with the imaging means A pixel number correcting means for weighting according to the distance between, and an estimating means for detecting the number of people in the region based on the number of pixels weighted by the pixel number correcting means. Special To.
[0008]
Further, the present invention is characterized in that, in the congestion monitoring system described above, the pixel number correcting unit divides the area into small areas of a predetermined size, and weighting is performed for each small area. To do.
[0009]
In addition, the present invention provides an imaging unit that captures an image of a predetermined region, a first storage unit that stores the image captured by the imaging unit, and a reference image of the region that serves as a reference captured in advance. The second storage means for storing, the value of each pixel constituting the image, and the value of each pixel constituting the reference image are compared with each other corresponding to each other, and a difference calculating means for calculating a difference between the pixel values Counting means for counting the number of pixels in which the difference in pixel values calculated by the difference calculation means exceeds a predetermined reference value; and a person virtually created using computer graphics Third storage means for storing the number of pixels and the number of persons calculated in a superimposed manner on a reference image, and the number of pixels counted by the counting means and the third storage means Remembered They based on the detected reference value, characterized in that it comprises an estimation means for detecting the number of people in the area.
[0010]
The present invention also includes a step of acquiring image data obtained by capturing an image of a predetermined area, a value of each pixel of the acquired image data, and a value of each pixel of a reference image of a reference area captured in advance. Comparing corresponding pixels with each other, calculating a pixel value difference, counting the number of pixels in which the calculated pixel value difference for each pixel exceeds a predetermined reference value, and The computer is caused to perform a step of weighting the number according to the position of each pixel and a step of detecting the number of persons in the region based on the weighted number of pixels.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a block diagram showing a configuration example of an embodiment of a congestion monitoring system according to the present invention. As shown in the figure, the present embodiment is configured by a camera 100 that captures a predetermined area such as a public facility, a road, or a park, a still image database (DB) 20 configured by a storage device, and a storage device. The background image DB 30 to be executed, the number of people estimation DB 40 constituted by the storage device, the weighting table 50 constituted by the storage device, the camera 100 and the still image DB 20, the background image DB 30, the number of people estimation DB 40, and the weighting table 50 are controlled. In addition, it is configured by a control unit 200 or the like that executes various processes to be described later.
[0012]
FIG. 2 is a functional block diagram of the present embodiment. The control unit 200 includes a frame division processing unit 210, a matching unit 211, a person number estimation processing unit 212, and the like. The frame division processing unit 210 divides the image data supplied from the camera 100 for each frame, extracts a still image, and stores it as a still image DB 20. The still image DB 20 includes a still image and a shooting time when the still image is shot. For example, pixel values are represented by 256 gray scales. Based on the still image read from the still image DB 20 and the background image read from the background image DB 30, the matching unit 211 compares the pixel values of corresponding pixels in units of pixels constituting each image. The difference is calculated. At this time, based on the mask area information (position coordinates corresponding to each pixel, weighting value, etc.), an effective image area is cut out from the imaging area, and the pixel value difference is equal to or greater than a predetermined reference value Is corrected (weighted). By performing this correction, for example, one pixel is detected for 10 pixels at pixels close to the shooting position, but one person is detected for 5 pixels at pixels far from the shooting position. The number of persons can be estimated in consideration of the fact that the size of a person photographed in inverse proportion to the distance from the camera becomes small. The number-of-people estimation processing unit 212 estimates the number of people in the area based on the number of pixels whose pixel value difference exceeds a predetermined reference value and the number-of-people estimation DB 40.
[0013]
FIG. 3 shows how the camera 100 captures a predetermined rectangular area ABCD (corresponding to a mask area in the captured image). The depth of this area is represented by W and the width is represented by L. The distance (height) between the position where the camera 100 is installed and the plane including the rectangular area ABCD is represented by H. FIG. 4 shows a rectangular area ABCD divided into lattice-shaped small areas 1 to 16. The small area can be, for example, a size of 40 cm (centimeter) × 40 cm where one person can enter.
[0014]
As the weight Qi for each small area in each area, a distance di (i is 1 to 16) from the position of the center of gravity of each small area to the position of the camera 100 is calculated, and the distance dmax from the camera 100 to the farthest small area is calculated. Then, the weight Qi is expressed by the following equation.
Qi = di ² / dmax ²
The weight Qi is calculated according to the distance of each small area, and is stored as mask area information in association with each small area of the cutout area of the captured image, as shown in FIG. The weighting by the weight Qi can be weighted so as to increase or decrease with respect to the number of pixels.
[0015]
As shown in FIG. 6, the data of the number of persons estimation DB is created when an image in which a person is placed on each small area of the rectangular area ABCD by using computer graphics is created and no person is placed. The difference between the value of each pixel of the background image and the value of each pixel of the image in which the person is arranged is obtained between the corresponding pixels, and the difference between the pixel values is virtually equal to the number of pixels greater than or equal to a predetermined reference value. Is stored in association with the number of persons placed in the. Here, one person or one column is erased in order from the person far from the camera 100, and the number of persons arranged and the number of pixels are associated with each other and stored as data in the person estimation DB 40.
[0016]
The size of the virtually created person is adjusted according to the distance from the camera 100. In addition, since the number of pixels varies depending on the height, the number of persons estimation DB 40 is created according to the height. For example, when the height of a virtually created person is 150 cm, it is estimated that 30 persons, and when the height is 170 cm, when the estimated number is 20 persons, the estimated number of persons can be 20 to 30 persons.
[0017]
Next, the operation of the present embodiment will be described with reference to the flowchart of FIG. First, in step S <b> 1, image data corresponding to an image captured by the camera 100 is supplied to the control unit 200. In step S <b> 2, the control unit 200 reads a corresponding background image from the background image DB 30 stored in the storage device 30. Next, in step S3, matching processing is executed. That is, the matching unit 211 of the control unit 200 compares the pixel values of the pixels constituting the image data captured by the camera 100 and the corresponding pixels of the background image.
[0018]
Next, in step S4, weighting set for each small region is performed on the pixel value of each pixel obtained by the matching processing executed in step S3. Next, in step S5, the number of people in the rectangular area ABCD is estimated from the weighted pixel value of each pixel and the data in the number-of-people estimation DB 40.
[0019]
Next, in step S6, the estimated number of persons is output to a display device or the like connected to the congestion monitoring system. Next, in step S7, it is determined whether or not an instruction to end the process is given, for example, by turning off the power. As a result, when it is determined that the end of the process is not instructed, the process returns to step S1 and the processes after step S1 are repeatedly executed. On the other hand, if it is determined that the end of the process has been instructed, the present process ends.
[0020]
Also, a predetermined warning can be given according to the estimated number of people. For example, the estimated number of people is compared with the previous data, and when the difference exceeds a predetermined value, it is detected that the number has suddenly increased, and an alarm is sounded or a message is displayed and managed. A warning can be made to the person. As a result, the manager can take measures to prevent troubles due to an increase in the number of people at the location where the image was taken, and remove the cause of the increase in the number of people.
[0021]
In the above-described embodiment, the number of persons estimation table 40 may be created according to the height of a person estimated to exist in the environment to be imaged. For example, when shooting a road near an adult-style venue on an adult-type day, the number-of-people estimation table 40 for a height of 150 cm (estimated 20-year-old woman) and a height of 170 cm (20-year-old male) (Estimate)) is prepared. Then, the number of people may be estimated by referring to these two number of people estimation tables 40. For example, when the number of pixels obtained by the matching process is A, the estimated number of people is 30 when the number of people estimation table 40 is 150 cm tall, and the number of people estimation table 40 is 170 cm tall. In the case where the estimated number of persons is 20, the estimated number of persons is “20 to 30 people” is output to the display device. For example, in the case of an elementary school yard, a database for estimating the number of persons for a height of 130 cm and a height of 150 cm can be created. Thus, by creating a database for estimating the number of persons according to the shooting environment, the estimated number of persons estimated according to the shooting location can be narrowed, and the reliability of the estimated number of persons can be improved.
[0022]
Further, the number of people estimation table 40 is different depending on the location where the camera is installed because, for example, a person is hidden behind a person in a case where a subject is photographed from a place where the depression angle is large and a depression angle is small. Is set.
[0023]
In addition, since the sunshine condition changes depending on time, weather, season, etc., a background image is prepared for each time, weather, season, and the number of people estimation table 40 is created for each time, weather, season, The number of people may be estimated by referring to the time image at the time of shooting, the weather, the background image according to the season, and the number of people estimation table 40. Thereby, the number of persons can be estimated in consideration of time, weather, and season.
[0024]
As described above, according to the present embodiment, the number of persons can be estimated in consideration of a deep environment, and the accuracy of the estimated number of persons can be improved. In addition, it is possible to estimate the number of people in consideration of people hiding behind the person. Furthermore, since the number of persons is estimated using computer graphics, it is possible to estimate the number of persons in consideration of the person hidden behind the person.
[0025]
In the above embodiment, the weighting is performed according to the square of the distance between each small region and the camera 100, but the present invention is not limited to this. For example, the upward direction of the image captured by the camera 100 may correspond to the depth distance. Further, assuming that the object in the y-axis direction of the image is far away, weighting proportional only to the y-axis direction may be performed.
[0026]
Further, in the above-described embodiment, when the photographed person image is not a single small area in the mask area but extends over a plurality of small areas, the mask area information of the small area at the foot of the person is set. The assigned weight Qi may be weighted for the whole pixel of the person, or may be weighted according to the weight Qi set for each small region.
[0027]
In the above-described embodiment, as shown in step S4 in FIG. 7, the number of people according to the weighted number of pixels is estimated by referring to the data stored in the number of people estimation DB 40. In other words, the data in which the number of pixels calculated using computer graphics is associated with the number of people is referred to, but the data to be referred to in order to estimate the number of people does not necessarily use computer graphics. It does not have to be the number of pixels calculated in the above. For example, the number of people may be estimated by simply referring to a table in which the number of pixels and the number of people are associated. Further, it may be calculated based on an equation for calculating the number of people from the number of pixels.
[0028]
Further, the program for realizing the functions of the frame division processing unit 210, the matching unit 211, and the number of people estimation processing unit 212 in FIG. Congestion monitoring may be performed by reading the system and executing it. The “computer system” here includes an OS and hardware such as peripheral devices.
[0029]
Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.
[0030]
The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes design and the like within a scope not departing from the gist of the present invention.
[0031]
【The invention's effect】
As described above, according to the congestion monitoring system and the congestion monitoring program according to the present invention, the number of people in a predetermined area such as public facilities, roads, parks, etc. is hidden in a deep environment or behind people. The number of persons can be estimated in consideration of the number of persons, and thereby the accuracy of the estimated number of persons can be improved.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration example of an embodiment of a congestion monitoring system according to the present invention.
FIG. 2 is a functional block diagram of the embodiment of FIG.
FIG. 3 is a diagram illustrating a state in which a camera captures a predetermined rectangular area ABCD (mask area).
FIG. 4 is a view showing lattice-shaped small areas 1 to 16 of a rectangular area ABCD.
FIG. 5 is a diagram for explaining a weight Qi.
FIG. 6 is a diagram showing an image in which a person is virtually arranged on each small area of a rectangular area ABCD.
FIG. 7 is a flowchart for explaining the operation of this embodiment;
[Explanation of symbols]
20 Still image DB 30 Background image DB
40 Number estimation DB 50 Weighting table 100 Camera 200 Control unit

Claims (5)

Imaging means for capturing an image of a predetermined area;
First storage means for storing the image picked up by the image pickup means;
Second storage means for storing a reference image of the region that is a reference imaged in advance;
A difference calculating means for comparing a value of each pixel constituting the image with a corresponding value of each pixel constituting the reference image and calculating a difference between the pixel values;
Counting means for counting the number of pixels in which the difference in pixel value for each pixel calculated by the difference calculating means exceeds a predetermined reference value;
Pixel number correcting means for weighting the number of pixels counted by the counting means in accordance with the position of each pixel;
Estimating means for detecting the number of people in the region based on the number of pixels weighted by the pixel number correcting means;
A congestion monitoring system comprising: Imaging means for capturing an image of a predetermined area;
First storage means for storing the image picked up by the image pickup means;
Second storage means for storing a reference image of the region that is a reference imaged in advance;
A difference calculating means for comparing a value of each pixel constituting the image with a corresponding value of each pixel constituting the reference image and calculating a difference between the pixel values;
Counting means for counting the number of pixels in which the difference in pixel value for each pixel calculated by the difference calculating means exceeds a predetermined reference value;
A number-of-pixels correcting unit that weights the number of pixels counted by the counting unit according to a distance from the imaging unit;
Estimating means for detecting the number of people in the region based on the number of pixels weighted by the pixel number correcting means;
A congestion monitoring system comprising: 3. The congestion monitoring according to claim 1, wherein the pixel number correcting unit divides the region into small regions of a predetermined size, and weighting is performed for each of the small regions. system. Imaging means for capturing an image of a predetermined area;
First storage means for storing the image picked up by the image pickup means;
Second storage means for storing a reference image of the region that is a reference imaged in advance;
A difference calculating means for comparing a value of each pixel constituting the image with a corresponding value of each pixel constituting the reference image and calculating a difference between the pixel values;
Counting means for counting the number of pixels in which the difference in pixel value for each pixel calculated by the difference calculating means exceeds a predetermined reference value;
A third storage means for storing the number of pixels and the number of people calculated by superimposing a person virtually created using computer graphics on the reference image;
Estimating means for detecting the number of persons in the region based on the number of pixels counted by the counting means and the detection reference value stored in the third storage means;
A congestion monitoring system comprising: Obtaining image data obtained by capturing an image of a predetermined area;
Comparing the value of each pixel of the acquired image data with the value of each pixel of the reference image of the reference imaged area that has been previously imaged between corresponding pixels, and calculating a difference in pixel values;
Counting the number of pixels for which the calculated pixel value difference for each pixel exceeds a predetermined reference value;
Weighting the counted number of pixels according to the position of each pixel;
Detecting the number of people in the region based on the weighted number of pixels;
A congestion monitoring program for causing a computer to execute.

Images