Monitoring control device, monitoring method, and computer program

Abstract

To accurately calculate the number of detection target objects present within a specified area based on images captured by an imaging device.SOLUTION: A monitoring control device 20 is configured to: calculate a first estimated number of detection target objects within a specified area by counting the number of detection target objects present in the specified area among detection target objects detected in an image captured by an imaging device 10; track the detection target objects detected in the image to count the number of detection target objects that enter the area and the number of detection target objects that have exited the area; calculate a second estimated number of detection target objects within the area based on a difference between the entry count number and the exit count number; and if the first estimated number and the second estimated number differ, adopt the larger number as the number of detection target objects in the area.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a monitoring control device, a monitoring method, and a computer program. [Background technology]

[0002] Patent Document 1 discloses a technology for detecting and tracking people in an image captured by a camera, thereby counting the number of people who pass through a detection line. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-201326 Summary of the Invention [Problem to be solved by the invention]

[0004] However, with the conventional technology described above, for example, if multiple people overlap and pass through the detection line, only the person in the foreground in the image may be counted, which could result in a low count.

[0005] The present invention has been made in response to such problems, and aims to accurately calculate the number of detection targets present in a specified area based on an image captured by an imaging device. [Means for solving the problem]

[0006] In order to solve the above problem, a monitoring control device according to one embodiment of the present invention is a monitoring control device that is configured to calculate the number of detection targets present in a specified area based on an image captured by an imaging device, and is configured to calculate a first estimated number of detection targets in the area by counting the number of detection targets present in the area among the detection targets detected in the image, count the number of detection targets that have entered the area and the number of detection targets that have exited the area as an entry count and an exit count, respectively, by tracking the detection targets detected in the image, calculate a second estimated number of detection targets in the area based on the difference between the entry count and the exit count, and when the first estimated number and the second estimated number differ, use the larger one as the number of detection targets in the area.

[0007] Furthermore, a monitoring method according to one aspect of the present invention is a monitoring method using a monitoring control device configured to calculate the number of detection targets present in a specified area based on an image captured by an imaging device, and calculates a first estimated number of detection targets in the area by counting the number of detection targets present in the area among the detection targets detected in the image, and counts the number of detection targets that have entered the area and the number of detection targets that have exited the area as an entry count and an exit count, respectively, by tracking the detection targets detected in the image, and calculates a second estimated number of detection targets in the area based on the difference between the entry count and the exit count, and when the first estimated number and the second estimated number differ, the larger one is used as the number of detection targets in the area.

[0008] Furthermore, a computer program according to one aspect of the present invention is a computer program for a monitoring control device that calculates the number of detection targets present in a specified area based on an image captured by an imaging device, and causes the monitoring control device to calculate a first estimated number of detection targets in the area by counting the number of detection targets present in the area among the detection targets detected in the image, count the number of detection targets that have entered the area and the number of detection targets that have exited the area as an entry count and an exit count, respectively, by tracking the detection targets detected in the image, calculate a second estimated number of detection targets in the area based on the difference between the entry count and the exit count, and when the first estimated number and the second estimated number differ, use the larger one as the number of detection targets in the area. [Effects of the Invention]

[0009] According to these aspects of the present invention, the number of detection targets within an area can be estimated using two different methods based on images captured by an imaging device, thereby enabling the number of detection targets present within an area to be calculated with high accuracy. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic configuration diagram of a monitoring device according to an embodiment of the present invention; [Figure 2] FIG. 1 is a diagram showing a surveillance image showing the passenger area inside the bus and the boarding and disembarking area outside the bus across the door. [Figure 3] 3 is a flowchart illustrating monitoring control according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0011] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, like components are designated by like reference numerals.

[0012] FIG. 1 is a schematic diagram of a monitoring device 1 according to an embodiment of the present invention.

[0013] The monitoring device 1 includes an imaging device 10 and a monitoring control device 20, and calculates the number of detection targets present in a predetermined area based on imaging data generated by the imaging device 10.

[0014] The imaging device 10 generates imaging data that captures a predetermined area to be monitored. In this embodiment, the imaging device 10 includes one or more cameras 11 for capturing images of the passenger area inside the bus. The cameras 11 capture images of at least the passenger area inside the bus at a predetermined frame rate (for example, 10 Hz to 40 Hz) and generate monitoring images (imaging data) that capture at least the passenger area inside the bus. Each time the cameras 11 generate a monitoring image, they transmit the generated monitoring image to the monitoring control device 20.

[0015] In this embodiment, the specified area is explained using an example in which the specified area is the passenger area on a bus, but of course the specified area is not limited to the passenger area on a bus, and can be, for example, part or all of an area inside a room, a venue, a train, etc.

[0016] The monitoring and controlling device 20 is an ECU (Electronic Control Unit) including a communication unit 21, a storage unit 22, and a processing unit .

[0017] The communication unit 21 includes an interface circuit for connecting the monitoring control device 20 to the internal network 2. The communication unit 21 supplies the monitoring image received from the camera 11 to the processing unit 23.

[0018] The storage unit 22 has a storage medium such as a hard disk drive (HDD), a solid disk drive (SSD), or a semiconductor memory, and stores various computer programs, data, and the like used for processing by the processing unit 23.

[0019] The processing unit 23 has one or more central processing units (CPUs) and their peripheral circuits, and executes various computer programs stored in the storage unit 22. The processing unit 23 is, for example, a processor. The processing unit 23 may further include other arithmetic circuits such as a logic unit, a numerical calculation unit, or a graphics processing unit. By executing processing in accordance with the computer programs, the processing unit 23 functions as a target detection unit 31, a first counting unit 32, a second counting unit 33, a target number determination unit 34, and a count number correction unit 35, and operates as functional units (modules) that realize predetermined functions. In the following description, when processing is described using each of the functional units 31 to 35 as the subject, it indicates that the processing unit 23 is executing a program that realizes each of the functional units 31 to 35.

[0020] The following describes the specific processing carried out by the monitoring control device 20. That is, the details of the functional units 31 to 35 realized by the processing unit 23 executing processing in accordance with a computer program will be described.

[0021] The target detection unit 31 detects a predetermined target appearing in the monitoring image as a detection target based on the monitoring image received from the camera 11. In this embodiment, the target detection unit 31 detects a person appearing in the monitoring image as a detection target by sequentially inputting the monitoring image to a classifier. The classifier may be, for example, a convolutional neural network (CNN) having multiple convolution layers connected in series from the input side to the output side. Note that the method for detecting targets is not limited to this method, and various known methods may be used for detection.

[0022] The first counting unit 32 and the second counting unit 33 count the number of detection targets present in a predetermined area using different methods. In this embodiment, the first counting unit 32 and the second counting unit 33 each count the number of people in the passenger area of ​​the bus.

[0023] Specifically, the first counting unit 32 simply counts the number of people present in the passenger area of ​​the bus among the people detected in the monitoring image, and calculates the number as the first count number.

[0024] On the other hand, the second counting unit 33 tracks people detected in the surveillance image in chronological order, counting the number of people who enter the passenger area inside the bus (entry count number) and the number of people who exit the passenger area (exit count number), and calculates the second count number by subtracting the exit count number from the entry count number.

[0025] Figure 2 shows a surveillance image of the passenger area inside the bus (unhatched area) and the boarding and disembarking area outside the bus separated by the door (hatched area), taken when the bus is stopped and the door is open.

[0026] 2, for example, when a person detected in a surveillance image crosses a virtual detection line drawn at an entrance / exit part of the passenger area of ​​the bus (for example, a door) from the boarding / disembarking area side toward the passenger area side, the second counting unit 33 determines that the person has entered the passenger area of ​​the bus. Conversely, when a person detected in a surveillance image crosses the detection line from the passenger area side toward the boarding / disembarking area side, the second counting unit 33 determines that the person has exited the passenger area of ​​the bus.

[0027] 2 shows one detection line, the number of detection lines is not limited to one and there may be multiple detection lines. Also, the detection lines may be used to create regions having shapes such as circles and squares.

[0028] The target number determination unit 34 determines the number of people present in the passenger area of ​​the bus based on the first count number and the second count number. When the first count number and the second count number are the same value, the target number determination unit 34 adopts that value as the number of people present in the passenger area of ​​the bus.

[0029] On the other hand, when the first count number and the second count number are different values, the target number determination unit 34 uses the larger one as the number of people present in the passenger area inside the bus. The reason for this will be explained below.

[0030] The first count number differs from the number of people actually present in the passenger area of ​​the bus when, for example, a person is partially or completely hidden in the passenger area, for example, by being overlapped by other people. In other words, the first count number tends to be counted as fewer than the number of people actually present in the passenger area of ​​the bus. This tendency is particularly pronounced when there are many targets in a relatively small area, such as the passenger area of ​​a bus.

[0031] Therefore, when the first count number is smaller than the second count number, the second count number is used as the number of targets present in a specified area, thereby enabling accurate calculation of the number of targets when a target is present in a specified area while being hidden by other targets.

[0032] On the other hand, cases in which the second count number differs from the number of people actually present in the passenger area on the bus may include when a person crosses the detection line and enters the area while being partially or completely hidden, for example, by overlapping with other people, or conversely, when a person exits the area.

[0033] In the former case, the number of entry counts will be less than the number of people who actually entered the area, so the second count will be less than the number of people actually present in the passenger area on the bus.In the latter case, the number of exit counts will be less than the number of people actually exiting the area, so the second count will be more than the number of people actually present in the passenger area on the bus.

[0034] Therefore, the second count number may be less than or more than the number of people actually present in the passenger area of ​​the bus. However, by adopting the first count number as the number of people present in the passenger area of ​​the bus when the second count number is smaller than the first count number, it is possible to accurately calculate the number of targets at least when a target crosses the detection line and enters the area while being hidden by another target.

[0035] In this way, when the first count number and the second count number are different values, the larger one is used as the number of people present in the passenger area of ​​the bus because this allows for accurate calculation of the number of targets when a target is present in a specified area while hidden by other targets, and also allows for accurate calculation of the number of targets when a target crosses the detection line and enters an area while hidden by other targets.

[0036] If the first count number and the second count number are different values, the count number correction unit 35 corrects the smaller count number to the larger count number.

[0037] For example, if the first count number is larger, the count number correction unit 35 corrects the entry count number or the exit count number so that the second count number becomes the first count number. In this embodiment, if the first count number is larger, the count number correction unit 35 corrects the entry count number by increasing it so that the second count number becomes the first count number. This is because, if the first count number is larger, it can be assumed that the first count number is larger because the target to be detected crossed the detection line while hidden by another target and entered the area.

[0038] On the other hand, if the second count number is larger, the count number correction unit 35 corrects the first count number to the second count number.

[0039] According to this embodiment, when the first count number and the second count number are different values, one of the count numbers is corrected each time, thereby suppressing a decrease in the accuracy of calculating the count number by the first counting unit 32 and the second counting unit 33, and ultimately suppressing a decrease in the accuracy of calculating the number of objects to be detected present in a specified area.

[0040] FIG. 3 is a flowchart illustrating the monitoring control according to this embodiment, which is performed by the monitoring control device 20.

[0041] In step S1, based on the monitoring image received from the camera 11, the monitoring control device 20 detects a person (detection target) appearing in the monitoring image.

[0042] In step S2, the monitoring control device 20 simply counts the number of people present in the passenger area of ​​the bus among those detected in the monitoring image, and calculates this as a first count number. The monitoring control device 20 also tracks the people detected in the monitoring image in chronological order to count the number of people who enter the passenger area of ​​the bus (entry count number) and the number of people who exit the passenger area (exit count number), and calculates a second count number by subtracting the exit count number from the entry count number.

[0043] In step S3, the monitoring control device 20 determines whether the first count number and the second count number are the same value. If the first count number and the second count number are the same value, the monitoring control device 20 proceeds to processing in step S4. On the other hand, if the first count number and the second count number are different values, the monitoring control device 20 proceeds to processing in step S5.

[0044] In step S4, since the first count number and the second count number are the same value, the monitor control device 20 determines that value as the number of people present in the passenger area inside the bus.

[0045] In step S5, since the first count number and the second count number are different values, the monitoring control device 20 adopts the larger count number as positive and sets the larger value as the number of people present in the passenger area of ​​the bus.

[0046] In step S6, the monitoring control device 20 corrects the smaller of the first and second count numbers to the larger count number. In this embodiment, as described above, if the first count number is larger, the monitoring control device 20 corrects the intrusion count number by increasing it so that the second count number becomes the first count number. If the second count number is larger, the monitoring control device 20 corrects the first count number to the second count number.

[0047] The monitoring control device 20 according to the present embodiment described above is configured to calculate the number of detection targets present in a predetermined area based on an image captured by the imaging device 10. Specifically, the monitoring control device 20 is configured to calculate a first count number (first estimated number) of detection targets in the area by counting the number of detection targets present in the area among the detection targets detected in the image, count the number of detection targets that have entered the area and the number of detection targets that have exited the area as the entry count number and the exit count number, respectively, by tracking the detection targets detected in the image, calculate a second count number (second estimated number) of detection targets in the area based on the difference between the entry count number and the exit count number, and, when the first count number and the second count number differ, use the larger one as the number of detection targets in the area.

[0048] The first count number generally tends to be counted as fewer than the number of people actually present in the specified area. Therefore, when the first count number is smaller than the second count number, by adopting the second count number as the number of targets present in the specified area, the number of targets can be accurately calculated when a target is present in the specified area while being hidden by other targets. Furthermore, when a target crosses the detection line while being hidden by other targets and enters the area, the second count number will be counted as fewer than the number of people actually present in the specified area. Therefore, when the second count number is smaller than the first count number, by adopting the first count number as the number of people present in the specified area, the number of targets can be accurately calculated at least when a target crosses the detection line while being hidden by other targets and enters the area.

[0049] In addition, the monitoring control device 20 according to this embodiment is configured to increase the entry count number so that when the first count number (first estimated number) is larger, the second count number (second estimated number) becomes the first count number.

[0050] When the first count number is larger, it can be assumed that the first count number is larger because the target to be detected crossed the detection line and entered the area while hidden by another target. Therefore, when the first count number is larger, the entry count number is increased so that the second count number becomes the first count number, thereby suppressing a decrease in the calculation accuracy of the second count number.

[0051] Furthermore, the monitoring control device 20 according to this embodiment is configured to correct the first count number (first estimated number) to the second count number when the second count number (second estimated number) is larger.

[0052] This makes it possible to prevent a decrease in the calculation accuracy of the first count number.

[0053] Although the embodiments of the present invention have been described above, the above embodiments merely illustrate some of the application examples of the present invention, and it is not intended that the technical scope of the present invention be limited to the specific configurations of the above embodiments.

[0054] For example, in the above embodiment, the completion control device 20 may further determine whether the area is over capacity or how crowded it is based on the number of detected objects present within a specified area, and may issue some kind of alert to people attempting to enter the area based on the results of this determination.

[0055] Also, for example, in the above embodiment, the computer program executed in the monitoring and control device 20 may be provided in a form recorded on a computer-readable portable recording medium such as a semiconductor memory, a magnetic recording medium, or an optical recording medium. [Explanation of symbols]

[0056] 10. Imaging device 20 Monitoring and control device

Claims

1. A monitoring control device that calculates the number of detection targets present in a predetermined area based on an image captured by an imaging device, calculating a first estimated number of the detection targets in the area by counting the number of the detection targets present in the area among the detection targets detected in the image; By tracking the detection targets detected within the image, the number of detection targets that have entered the area and the number of detection targets that have exited the area are counted as an entry count and an exit count, respectively, and a second estimated number of the detection targets within the area is calculated based on the difference between the entry count and the exit count; When the first estimated number and the second estimated number differ, the larger one is adopted as the number of the detection targets in the area. Monitoring and control equipment.

2. When the first estimated number is larger, the entry count number is corrected so that the second estimated number becomes the first estimated number. The monitoring and control device according to claim 1 .

3. When the second estimated number is larger, the first estimated number is corrected to the second estimated number. The monitoring and control device according to claim 1 or 2.

4. A monitoring method using a monitoring control device configured to calculate the number of detection targets present in a predetermined area based on an image captured by an imaging device, calculating a first estimated number of the detection targets in the area by counting the number of the detection targets present in the area among the detection targets detected in the image; By tracking the detection targets detected within the image, the number of detection targets that have entered the area and the number of detection targets that have exited the area are counted as an entry count and an exit count, respectively, and a second estimated number of the detection targets within the area is calculated based on the difference between the entry count and the exit count; When the first estimated number and the second estimated number are different, the larger one is adopted as the number of the detection targets in the area. Monitoring method.

5. A computer program for a monitoring control device that calculates the number of detection targets present in a predetermined area based on an image captured by an imaging device, calculating a first estimated number of the detection targets in the area by counting the number of the detection targets present in the area among the detection targets detected in the image; By tracking the detection targets detected within the image, the number of detection targets that have entered the area and the number of detection targets that have exited the area are counted as an entry count and an exit count, respectively, and a second estimated number of the detection targets within the area is calculated based on the difference between the entry count and the exit count; causing the monitoring control device to adopt, when the first estimated number and the second estimated number are different, the larger one as the number of the detection targets in the area; Computer program.

Images