JP2022159930A - Monitoring camera installation support system, monitoring camera installation support device, and monitoring camera installation support program - Google Patents

Abstract

To enable an installer to easily confirm a change in the accuracy of an analysis result due to adjustment of an installation situation when the installer adjusts the installation situation of a monitoring camera such as the installation position and the installation angle of the monitoring camera in a monitoring camera installation support system.SOLUTION: A superimposed image obtained by superimposing information on an analysis result of each frame image input from a monitoring camera on the frame image is generated, and a superimposed video 91 composed of these superimposed images is displayed together with camera installation information 92 (for example, installation position and installation angle of the monitoring camera) associated with this superimposed video 91. As a result, when an installation operator adjusts the installation condition of the monitoring camera, such as the installation position and installation angle of the monitoring camera, the installation operator can easily check the change in the accuracy of the analysis result due to the adjustment of the installation status of the monitoring camera on the basis of the camera installation information 92 indicating the installation condition and analysis result information (human detection box 91a and face detection box 91b) superimposed on each frame image of the superimposed video 91.SELECTED DRAWING: Figure 17

Description

The present invention relates to a surveillance camera installation assistance system, a surveillance camera installation assistance device, and a surveillance camera installation assistance program.

Conventionally, surveillance cameras, such as so-called AI (Artificial Intelligence) cameras, have been placed in stores, etc., and analysis of frame images taken by the surveillance cameras (analysis of learned object detection for objects such as people reflected in the frame images) Devices and systems that perform detection using a neural network, recognition using a trained neural network for object recognition, etc. are known (see, for example, Patent Document 1).

JP 2018-093283 A

However, in a device or system for analyzing frame images captured by a conventional surveillance camera, such as that described in Patent Document 1, due to the influence of the installation position and installation angle of the surveillance camera, the light source around the surveillance camera, etc., Accuracy of analysis results (for example, accuracy of detection and recognition of objects such as people) may not be sufficiently obtained. Therefore, in order to ensure the accuracy of the analysis results in this type of device or system, when the installer adjusts (changes) the installation position or installation angle of the surveillance camera at the installation site of the surveillance camera, the analysis result It is necessary to be able to easily confirm changes in accuracy (for example, accuracy of detection and recognition of objects such as people).

The present invention is intended to solve the above problems, and when an installer adjusts (changes) the installation status of a surveillance camera such as the installation position and installation angle of the surveillance camera at the installation site of the surveillance camera, the installation situation can be adjusted. It is an object of the present invention to provide a monitoring camera installation support system, a monitoring camera installation support device, and a monitoring camera installation support program that enable easy confirmation of changes in the accuracy of analysis results obtained by monitoring.

In order to solve the above problems, a monitoring camera installation support system according to a first aspect of the present invention is a monitoring camera installation support system for supporting installation of a monitoring camera that captures images of a predetermined shooting area, wherein: Installation information acquisition means for acquiring camera installation information indicating the installation status at the time of shooting; image analysis means for analyzing each frame image input from the surveillance camera in the installation status; and image analysis means for each frame image. analysis result superimposing means for generating a superimposed image obtained by superimposing the information of the analysis result of each frame image by superimposing; video storage means; installation information storage means for storing the camera installation information obtained by the installation information acquisition means at the time of shooting of the monitoring camera in association with the superimposed video stored in the superimposed video storage means; and video display means for displaying the superimposed video stored in the superimposed video storage means together with the camera installation information linked to the superimposed video stored in the installation information storage means.

In this monitor camera installation support system, it is preferable that the image display means display superimposed images in a plurality of installation situations of the monitor cameras side by side.

In this surveillance camera installation support system, the camera installation information stored in the installation information storage means in another surveillance camera installation support system installed at a location different from the surveillance camera installation support system is selected. camera installation information retrieval means for retrieving camera installation information corresponding to an installation status similar to the installation status of the surveillance camera, wherein the video display means links the camera installation information retrieved by the camera installation information retrieval means It is desirable to display the superimposed image thus retrieved together with the retrieved camera installation information.

In this surveillance camera installation support system, it is preferable that the video display means display, side by side, each of the superimposed images associated with the plurality of camera installation information retrieved by the camera installation information retrieval means.

In this surveillance camera installation support system, it is preferable that the camera installation information includes information on the installation position and installation angle of the surveillance camera.

In this surveillance camera installation support system, the surveillance camera installation support system includes a mobile terminal, and the installation information obtaining means obtains information from a sensor including a camera mounted on the mobile terminal to obtain the monitoring information. Information on the installation position and installation angle of the camera may be acquired.

In this surveillance camera installation support system, the installation information acquisition means obtains the installation position and installation angle of the surveillance camera based on the image of the two-dimensional bar code board installed in the shooting area taken by the surveillance camera. information may be acquired.

In this surveillance camera installation support system, the installation information acquiring means may estimate the installation height position and installation angle of the surveillance camera based on the image of each mark of the two-dimensional bar code.

A surveillance camera installation assistance device according to a second aspect of the present invention is a surveillance camera installation assistance device for assisting installation of a surveillance camera for photographing a predetermined photographing area, in which each frame image input from the surveillance camera is analyzed. analysis result superimposition means for generating a superimposed image in which information of the analysis result of each frame image by the image analysis means is superimposed on each frame image; and each superimposition generated by the analysis result superimposition means superimposed video storage means for storing a superimposed video composed of images; and installation information storage means for storing camera installation information at the time of shooting of the surveillance camera in association with the superimposed video stored in the superimposed video storage means. and video output means for outputting the superimposed video stored in the superimposed video storage means to display means together with the camera installation information linked to the superimposed video stored in the installation information storage means.

A surveillance camera installation assistance program according to a third aspect of the present invention is a surveillance camera installation assistance program for assisting installation of a surveillance camera for photographing a predetermined photographing area, comprising: installation information acquisition means for acquiring camera installation information indicating installation status; image analysis means for analyzing each frame image input from the surveillance camera in the installation status; Analysis result superimposing means for generating a superimposed image obtained by superimposing information on the analysis result of the frame image, and superimposed video storage means for storing a superimposed video composed of each superimposed image generated by the analysis result superimposing means in the installation situation. and installation information storage means for storing the camera installation information obtained by the installation information acquisition means at the time of shooting of the monitoring camera in association with the superimposed video stored in the superimposed video storage means, and the superimposed video storage. The superimposed video stored in the means is functioned as video display means for displaying the camera installation information linked to the superimposed video stored in the installation information storage means.

According to the surveillance camera installation assistance system according to the first aspect of the present invention and the surveillance camera installation assistance program according to the third aspect of the present invention, the information of the analysis result of each frame image is superimposed on each frame image input from the surveillance camera. A superimposed image is generated, and a superimposed image composed of these superimposed images is displayed together with camera installation information linked to this superimposed image (for example, information on the installation position and installation angle of the surveillance camera). . As a result, when an installation worker (of a surveillance camera) adjusts (changes) the installation status of a surveillance camera, such as the installation position and installation angle of the surveillance camera at the installation site of the surveillance camera, the camera that indicates the installation status Along with the installation information, look at the analysis result information superimposed on each frame image of the superimposed video (for example, the detection results and recognition results of people, etc.), and adjust the installation status of the surveillance cameras to determine changes in the accuracy of the analysis results. can be easily verified.

According to the monitoring camera installation support device according to the second aspect of the present invention, a superimposed image is generated by superimposing the information of the analysis result of each frame image on each frame image input from the monitoring camera, and from these superimposed images, The constructed superimposed video is output to the display means and displayed together with the camera installation information linked to the superimposed video (for example, information on the installation position and installation angle of the surveillance camera). As a result, when an installation worker (of a surveillance camera) adjusts (changes) the installation status of a surveillance camera, such as the installation position and installation angle of the surveillance camera at the installation site of the surveillance camera, the camera that indicates the installation status Along with installation information, analysis result information superimposed on each frame image of the superimposed video (for example, detection results and recognition results of people, etc.) can be viewed on the display means. You can easily see changes in the accuracy of your results.

1 is a block configuration diagram showing a schematic configuration of a monitoring camera installation support system according to an embodiment of the present invention; FIG. FIG. 2 is a block diagram showing a schematic hardware configuration of an analysis box in FIG. 1; FIG. 2 is a block diagram showing a schematic hardware configuration of the smartphone in FIG. 1; FIG. 2 is a block diagram showing a schematic hardware configuration of the tablet in FIG. 1; FIG. 2 is a functional block configuration diagram of the monitoring camera, analysis box, smartphone, and tablet in FIG. 1; FIG. 6 is a first explanatory diagram of a superimposed video storage sequence performed by the video acquisition unit, the image analysis unit, and the analysis result superimposition unit in FIG. 5; FIG. 11 is a second explanatory diagram of the superimposed video storage sequence; FIG. 6 is an explanatory diagram of a superimposed image search/display sequence performed by an image acquisition unit, an analysis result superimposition unit, and an image confirmation unit in FIG. 5; FIG. 10 is an explanatory diagram of an example of a method of acquiring an installation position and an installation angle of a monitoring camera when the smartphone is provided with an installation information acquisition unit; FIG. 4 is an explanatory diagram of an example of a method of acquiring a three-dimensional position of lighting when the smartphone is provided with an installation information acquiring unit; FIG. 5 is an explanatory diagram of an example of a method of acquiring information such as lighting type, intensity, and color when the smartphone is provided with an installation information acquiring unit; (a) is a diagram showing a photographed image of a board on which a two-dimensional barcode is printed, and (b) is an explanatory diagram of XYZ coordinates of installation positions of monitoring cameras. The figure which shows the superimposed image generation search screen displayed on a tablet. FIG. 10 is a view showing a superimposed video recording screen displayed on the tablet when recording a superimposed video; FIG. 10 is a view showing a superimposed video recording completion screen displayed on the tablet when recording of a superimposed video is completed; 14 is a view showing an in-store camera installation state selection screen displayed on the tablet when the similar camera image display button of the main store in FIG. 13 is pressed; FIG. FIG. 17 is a view showing an in-store camera video search result display screen displayed on the tablet when the installation state in FIG. 16 is selected; The figure which shows the other store camera installation situation selection screen displayed on a tablet, when the similar camera image display button of the other store of FIG. 13 is pushed. FIG. 19 is a diagram showing a display screen of other store camera video search results displayed on the tablet when the installation state in FIG. 18 is selected;

Hereinafter, a monitoring camera installation support system, a monitoring camera installation support device, and a monitoring camera installation support program according to embodiments embodying the present invention will be described with reference to the drawings. FIG. 1 is a block configuration diagram showing a schematic configuration of a monitoring camera installation support system 1 according to this embodiment. In the present embodiment, a plurality of monitoring cameras 2, which are network cameras (IP (Internet Protocol) cameras) for monitoring a predetermined shooting area, and an analysis box 3 to which these monitoring cameras 2 are connected. (corresponding to a "surveillance camera installation support device") is arranged in a store S such as a chain store. As shown in FIG. 1, the surveillance camera installation support system 1 includes a smartphone (hereinafter abbreviated as "smartphone") 4 (hereinafter abbreviated as "smartphone") in a store S in addition to the above surveillance camera 2 and analysis box 3. (equivalent to a “mobile terminal”), a tablet 5 , a WiFi AP (WiFi Access Point) 6 , a hub 7 and a router 8 .

The analysis box 3 is connected to a smartphone 4 and a tablet 5 via a WiFi AP 6 and a hub 7, and is connected to a plurality of surveillance cameras 2 via a LAN (Local Area Network) 20 and a hub 7, Images input from each of these surveillance cameras 2 are analyzed. Specifically, the analysis box 3 performs object detection processing (e.g., person (customer) or face detection processing) on images input from each of the monitoring cameras 2, and an object image detected by this object detection processing. object recognition processing (for example, recognition processing of attributes such as gender and age of a person (customer)). In addition, the analysis box 3 adds information on the analysis results of each frame image (for example, detection results and recognition results of people, etc.) to each frame image input from the monitoring camera 2 when the monitoring camera 2 is installed or adjusted. A superimposed image is generated, and a superimposed image composed of these superimposed images is displayed on the tablet 5 together with camera installation information (for example, information on the installation position and installation angle of the surveillance camera 2) linked to the superimposed image. output to and display. When the installation information acquisition unit (see FIG. 5) is provided in the analysis box 3, the analysis box 3 and the tablet 5 correspond to the "computer" in the claims, and when the installation information acquisition unit is provided in the smartphone 4 , the analysis box 3, the smartphone 4, and the tablet 5 correspond to the "computer" in the claims, and when the installation information acquisition unit is provided in the surveillance camera 2 to be installed (or adjusted), The box 3 and the tablet 5 correspond to the "computer" in the claims.

The monitoring camera installation support system 1 also includes an AI analysis server 9 on the cloud C and a management server 10 . The AI analysis server 9 analyzes, for example, the behavior of people in each store based on the object detection results and object recognition results from the analysis box 3, and uses the information of the analysis results for various purposes such as marketing and crime prevention. The application converts it into easy-to-use data and outputs it.

The management server 10 described above manages a large number of analysis boxes 3 arranged in each store and the monitoring cameras 2 connected to these analysis boxes 3 . Specifically, the management server 10 performs control such as installation of application packages to the analysis boxes 3 of each store, and starting and stopping of the monitoring cameras 2 connected to these analysis boxes 3 . This application package includes a trained neural network model for object detection (hereinafter referred to as "NN model for object detection") and a trained neural network model for object recognition (hereinafter referred to as "NN model for object recognition"). ) and a control script that describes how to use (order of processing) these neural network models (hereinafter referred to as “NN models”), and is a program for supporting installation of surveillance cameras. It is included.

Next, the hardware configuration of the analysis box 3 will be described with reference to FIG. The analysis box 3 includes a CPU 11 that controls the entire device and performs various calculations, a hard disk 12 that stores various data and programs, a RAM (Random Access Memory) 13, and an inference chip (hereinafter referred to as " 14a to 14h and a communication control IC 15. The CPU 11 is a general-purpose CPU or a CPU designed to enhance parallel processing performance in order to simultaneously process multiple video streams. The data stored in the hard disk 12 includes various DBs (data bases) such as a video DB 54 (see FIG. 5), which will be described later. The programs stored in the hard disk 12 include the analysis box OS program in addition to the above application package 17 (main part of the "monitoring camera installation support program" in the claims). In addition to the application package 17 described above, the "surveillance camera installation support program" in the claims includes an installation information acquisition application 32 for the smartphone 4 and a superimposed image generation confirmation program 47 for the tablet 5, which will be described later.

The above (inference) chips 14a to 14h are preferably processors (inference-dedicated chips) optimized for DNN (Deep Neural Networks) inference. ), or any other processor. Further, each of the chips 14a to 14h may be a device in which a plurality of chips (inference processors) are integrated (mounted) on one board computer.

As shown in FIG. 2, the above (inference) chips 14a-14h are connected to the CPU 11 by PCI Express or USB. The communication control IC 15 also has a LAN port 16 that is a port for connection to a LAN conforming to the Ethernet standard.

Next, the hardware configuration of the smartphone 4 and the installation information acquisition application 32 installed on the smartphone 4 will be described with reference to FIG. The smartphone 4 includes an SoC (System-on-a-Chip) 21, a built-in camera 24, a microphone 25, a touch panel display 26, a speaker 27, a memory 28 for storing various data and programs, and a communication unit 29. , a secondary battery 30 , and a charging terminal 31 . The SoC 21 includes a CPU 22 that controls the entire device and performs various calculations, and a GPU 23 that is used for inference processing of the NN model for object detection and the NN model for object recognition. Also, the touch panel display 26 allows the user (installer of the surveillance camera 2) to specify the position of the light source (illumination, etc.) around the surveillance camera 2 in the input image from the built-in camera 24 displayed on the touch panel display 26. Also used as a pointing device for

The programs stored in the memory 28 include an installation information acquisition application 32 downloaded from the management server 10 to the smartphone 4 and installed. In the example shown in FIG. 3 , the installation information acquisition application 32 includes a camera detection model 33 and an AR application 34 . The AR application 34 is a program for an AR engine using existing technology such as ARKit (a framework for AR-compatible applications for iPhone (registered trademark) and iPad (registered trademark)). However, as will be described later, when the user uses the UI of the smartphone 4 to manually input camera installation information indicating the installation status of the surveillance camera 2 at the time of shooting, the installation information acquisition application 32 performs the camera detection described above. It is not necessary to include the model 33 and the AR application 34, and it is only necessary to include a program for realizing a UI for inputting camera installation information.

The communication unit 29 has a communication IC and an antenna. The smartphone 4 is connected to the management server 10 via the communication unit 29 and the network. The secondary battery 30 is a battery such as a lithium ion battery that can be used repeatedly by charging, and stores electric power from a commercial power supply after being converted into direct-current power by an AC/DC converter. Supplied to each part of the smartphone 4.

Next, referring to FIG. 4, the hardware configuration of the tablet 5 will be described. The tablet 5 includes a built-in camera 40, an SoC (System-on-a-Chip) 41, a touch panel display 44, a speaker 45, a memory 46 for storing various data and programs, a communication unit 48, a secondary A battery 49 and a charging terminal 39 are provided. The SoC 41 includes a CPU 42 that controls the entire device and performs various calculations, and a GPU 43 that is used for inference processing of various learned DNN (Deep Neural Networks) models.

The programs stored in the memory 46 include a superimposed image generation confirmation program 47, which is a program for realizing the function of the image confirmation unit 63 described later with reference to FIG. This superimposed video generation confirmation program 47 has a function of inputting an instruction to generate a superimposed video in which the information of the analysis result is superimposed on the frame image input from each of the monitoring cameras 2, and the superimposed video generation process. A function to search and display superimposed images of surveillance cameras 2 whose installation status is similar to the installation status of the surveillance camera 2 to be installed this time, and from among the surveillance cameras 2 placed in other stores, This is a program for realizing a function of inputting an instruction to search and display a superimposed image of a surveillance camera 2 having an installation situation similar to that of the surveillance camera 2 to be installed this time.

The communication unit 48 described above includes a communication IC and an antenna. The tablet 5 is connected to the analysis box 3 and the management server 10 on the cloud C via the communication unit 48 and the network. Also, the secondary battery 49 is a battery such as a lithium ion battery that can be used repeatedly by charging, and stores the electric power from the commercial power source after being converted into DC power by the AC/DC converter. Supply to each part of the tablet 5 .

FIG. 5 shows an example of functional blocks of the monitoring camera 2, analysis box 3, smartphone 4, and tablet 5 described above. The analysis box 3 includes a video acquisition unit 53, an image analysis unit 55, and an analysis result superimposition unit 57 as main functional blocks. The analysis result superimposing unit 57 includes a superimposing unit 58 , an installation information linking unit 59 , and a query receiving unit 61 . The query reception unit 61 has a video output unit 62 . The video acquisition unit 53 , the image analysis unit 55 , the superimposition unit 58 , the installation information connection unit 59 , and the query reception unit 61 described above are functional blocks of the CPU 11 of the analysis box 3 . The CPU 11 of the analysis box 3 implements these functions using the application package 17 described above. The analysis box 3 also includes a video DB 54, an analysis result DB 56, and an installation information DB 60 on the hard disk 12 (see FIG. 2). The image analysis unit 55, the analysis result superimposition unit 57, the image DB 54, the installation information DB 60, and the image output unit 62 are respectively the image analysis means, the analysis result superimposition means, and the superimposed image storage. means, installation information storage means, and video output means.

The monitoring camera 2 described above includes an imaging unit 50 . Further, in the example shown in FIG. 5, the smartphone 4 includes an installation information acquisition unit 52. However, as indicated by the dashed line in FIG. 5 , the monitoring camera 2 may include the installation information acquisition section 51 , and the analysis box 3 may include the installation information acquisition section 68 . The installation information acquisition unit 51, the installation information acquisition unit 52, or the installation information acquisition unit 68 described above corresponds to installation information acquisition means in the claims. The installation information acquisition unit 51, the installation information acquisition unit 52, or the installation information acquisition unit 68 acquires camera installation information indicating the installation status of the monitoring camera 2 at the time of shooting. The installation information acquisition unit 52 described above is a functional block of the SoC 21 (see FIG. 3) of the smartphone 4, and the installation information acquisition unit 68 is a functional block of the CPU 11 (see FIG. 2) of the analysis box 3. The installation information acquisition unit 51 described above is a functional block included in the CPU (not shown) of the surveillance camera 2 . In the following description, an example in which the smartphone 4 includes the installation information acquisition unit 52 will be mainly described.

The video acquisition unit 53 in the analysis box 3 described above is a function realized by a so-called VMS (Video Management System), and decodes the video stream input from the monitoring camera 2 via the hub 7, The frame image (data of) is stored in the video DB 54 of the hard disk 12 and is output to the image analysis section 55 via the bus in the analysis box 3 . The video acquisition unit 53 constitutes a VMS server together with the video DB 54 .

The image analysis unit 55 described above analyzes each frame image input from the surveillance camera 2 . The analysis result DB 56 stores (stores) information on the analysis result of each frame image by the image analysis unit 55 . The analysis result superimposing unit 57 generates a superimposed image by superimposing the information of the analysis result of each frame image by the image analysis unit 55 on each frame image input from the monitoring camera 2 . The video DB 54 stores superimposed video composed of superimposed images generated by the analysis result superimposing unit 57 in addition to the decoded frame images input from the surveillance camera 2 . More precisely, the superimposing unit 58 in the analysis result superimposing unit 57 adds the analysis result information of each frame image stored in the analysis result DB 56 to each frame image from the monitoring camera 2 stored in the video DB 54 . superimposed images are generated, and a superimposed video composed of each generated superimposed image is stored in the video DB 54 . In addition, the installation information linking unit 59 receives the camera installation information sent from the installation information acquiring unit 52 of the smartphone 4, together with the video start time of the superimposed image generated by the superimposing unit 58 (superimposed image generation start time), Write to the installation information DB 60 . As a result, the installation information DB 60 stores the camera installation information obtained by the installation information acquisition unit 52 on the smartphone 4 side at the time when the monitoring camera 2 takes the picture, in association with the superimposed video stored in the video DB 54. .

Further, the tablet 5 described above includes an image confirmation section 63 . The video confirmation section 63 includes a display section 64 , a condition input section 66 and a query issuing section 67 . The display unit 64 (corresponding to video display means in the claims) displays the superimposed video stored in the video DB 54 together with the camera installation information linked to the superimposed video stored in the installation information DB 60 .

As shown in FIG. 5, the surveillance camera installation support system 1 includes a camera installation information search section 65 (corresponding to camera installation information search means in claims). The camera installation information search unit 65 is composed of a condition input unit 66 and a query issuing unit 67 on the tablet 5 side, and a query receiving unit 61 on the analysis box 3 side.

The camera installation information search unit 65 searches for the installation status of the surveillance camera 2 to be installed this time from among the camera installation information stored in the installation information DB 60 of the analysis box 3 in FIG. It has a function to search for camera installation information corresponding to similar installation situations. In other words, the camera installation information retrieving unit 65 monitors the installation status similar to the installation status of the surveillance camera 2 to be installed this time from among the superimposed images based on the images of the surveillance cameras 2 installed in the same store S. It has a function of searching the superimposed video of the camera 2 . The camera installation information search unit 65 also retrieves the installation information DB 60 of another store (another monitor camera installation support system installed at a different location from the monitor camera installation support system 1 in FIGS. 1 and 5) via the network. It has a function of retrieving camera installation information corresponding to an installation situation similar to the installation situation of the surveillance camera 2 to be installed from among the camera installation information stored in the installation information DB 60 in the system 1 . In other words, the camera installation information search unit 65 selects a surveillance camera whose installation status is similar to that of the surveillance camera 2 to be installed this time, from among the superimposed images based on the images of the surveillance cameras 2 installed in other stores. 2 has a function of searching for superimposed images.

The display unit 64 displays the superimposed image associated with the camera installation information retrieved by the camera installation information retrieval unit 65 together with the retrieved camera installation information. Further, although the details will be described later, when there is a plurality of pieces of camera installation information retrieved by the camera installation information retrieval unit 65, the display unit 64 displays each of the superimposed images linked with the plurality of pieces of camera installation information. They are displayed side by side (see FIGS. 17 and 19).

Next, referring to FIGS. 6 and 7, the superimposed video storage sequence mainly performed by the video acquisition unit 53, the image analysis unit 55, and the analysis result superimposition unit 57 will be described. First, the analysis result superimposing unit 57 in the analysis box 3 inputs from the smartphone 4 the camera installation information indicating the installation status of the surveillance camera 2 at the time of shooting, which is acquired by the installation information acquiring unit 52 of the smartphone 4 in FIG. . This camera installation information is information indicating installation conditions such as the installation position and installation angle of the monitoring camera 2 at the time of shooting, the light source condition around the monitoring camera 2, and the detection location of the detection target. The installation information acquisition unit 52 may be realized by the user manually inputting the camera installation information using the UI of the smartphone 4 . When the superimposed image generation button 79 on the superimposed image generation search screen 78 (see FIG. 13) displayed on the touch panel display 44 of the tablet 5 is pressed by the user, the analysis result superimposition unit 57 displays the installation position of the monitoring camera 2, etc. is stored in the installation information DB 60 together with the video start time (superimposed image generation start time) (S1).

On the other hand, when the superimposed image generation button 79 (see FIG. 13) is pressed, the image acquisition unit 53 acquires an image (stream) from the surveillance camera 2 (S2), decodes it, and generates a frame after decoding. The image is transmitted to the image analysis unit 55 (S3), and the camera image composed of the frame image is stored in the image DB 54 (S4).

In addition, when the image analysis unit 55 acquires the frame image sent from the video acquisition unit 53 (VMS) in S3, the image analysis unit 55 inputs this frame image to various trained DNN models to obtain detection results of people, faces, etc. Also, information (inference information) of recognition results such as gender and age is acquired (S6). After obtaining the inference information, the image analysis unit 55 transmits the inference information and the frame image obtained in S5 to the analysis result superimposition unit 57 (S7).

When the analysis result superimposing unit 57 acquires the frame image and the inference information (inference result) transmitted from the image analysis unit 55 in S7 (S8), it generates a superimposed image in which the inference result is superimposed on each acquired frame image. (S9), a superimposed video is generated from these superimposed images, and transmitted to the video acquisition unit 53 (VMS) (S10). The video acquisition unit 53 saves this superimposed video in the video DB 54 (S11).

After the process of S10, the analysis result superimposing unit 57 calculates statistical information such as the number of people detected based on the inference result received from the image analysis unit 55, and stores the statistical information in the analysis result DB 56 (S12). This statistical information is statistical information such as the total number of people and faces detected in each superimposed video for a predetermined time period (for example, 1 to 2 minutes) stored in the video DB 54 and inference accuracy. It does not have to be automatically calculated by the superimposing unit 57, and may be manually input by the operator using the smartphone 4 or the tablet 5, for example.

Next, referring to FIG. 8, the superimposed video search/display sequence performed mainly by the video acquisition unit 53 and the analysis result superimposition unit 57 on the analysis box 3 side and the video confirmation unit 63 on the tablet 5 side will be described. do. The user selects a superimposed image of the surveillance camera 2 in an installation situation similar to the installation situation of the surveillance camera 2 during installation (or adjustment) work from the condition input section 66 (see FIG. 5) of the image confirmation section 63 on the tablet 5 side. When camera installation information such as a camera installation position is entered as a search key for a search, the query issuing unit 67 uses the entered camera installation information as a key to the analysis result superimposition unit 57 on the analysis box 3 side, A query is issued (S21). Note that the query issuing unit 67 selects, from among the superimposed images based on the images of the surveillance cameras 2 installed in the same store S, the surveillance cameras 2 whose installation status is similar to that of the surveillance camera 2 to be installed this time. When retrieving the superimposed video, a query is issued to the analysis result superimposing unit 57 of the analysis box 3 installed in the same store S. In addition, the query issuing unit 67 superimposes a monitoring camera 2 having an installation state similar to that of the monitoring camera 2 to be installed this time, from the superimposed video based on the video of the monitoring camera 2 installed in another store. When retrieving a video, a query is issued to the analysis result superimposing unit 57 of the analysis box 3 installed in another store.

When the query receiving unit 61 receives the query issued by the query issuing unit 67 on the tablet 5 side, the analysis result superimposing unit 57 on the analysis box 3 side selects the query from the camera installation information stored in the installation information DB 60 . (S22). Then, the video start time of the superimposed video linked to the camera installation information, which is the search result, is read from the installation information DB 60, and a superimposed video acquisition query with this video start time as a key is issued to the video acquisition unit 53. (S23). The video acquisition unit 53 searches the superimposed video stored in the video DB 54 for a superimposed video having a video start time corresponding to the video start time included in the superimposed video acquisition query (S24). The superimposed image is sent to the image confirmation section 63 on the tablet 5 side (S25).

After the end of the process of S23, the analysis result superimposing unit 57 acquires the statistical information of the superimposed video retrieved by the video acquiring unit 53 in S24 from the analysis result DB 56, and the acquired statistical information is Together with the camera installation information retrieved in S22, it is transmitted to the video confirmation unit 63 on the tablet 5 side (S26). The video confirmation unit 63 on the tablet 5 side receives the superimposed video, camera installation information, and statistical information transmitted from the analysis box 3 side in S25 and S26 (S27 and S28), and checks the received superimposed video and camera installation information. and statistical information are displayed on the display unit 64 (S29).

Next, with reference to FIG. 9, an example of a camera installation information acquisition method when the smartphone 4 is provided with the installation information acquisition unit 52 (see FIG. 5) will be described. Before executing the camera installation information acquisition process by this acquisition method, the installer of the surveillance camera 2 downloads the installation information acquisition application 32 (see FIG. 3) from the management server 10 (see FIG. 1) to the smartphone 4. and install it. Then, when installing or adjusting the monitoring camera 2, when the installation worker activates the installation information acquisition application 32 and points the smartphone 4 at the monitoring camera 2, the SoC 21 of the smartphone 4 performs the installation shown in FIG. After detecting the surveillance camera 2 using the camera detection model 33 included in the information acquisition application 32, the three-dimensional position of the surveillance camera 2 is acquired by the AR function of the AR application 34 (see FIG. 3). For example, using the built-in camera 24 (see FIG. 3), a GPS (Global Positioning System) sensor, an electronic compass, and an acceleration sensor of the smartphone 4, the three-dimensional position (installation position) and Get the installation angle. Specifically, in addition to the image 69 taken by the built-in camera 24 of the smartphone 4, the location information of the smartphone 4 captured by the GPS sensor (GPS module) and the image taken by the built-in camera 24 acquired using the electronic compass. The three-dimensional position (installation position) and installation angle of the monitoring camera 2 are acquired based on the direction and the tilt of the smartphone 4 acquired using the acceleration sensor. In FIG. 9, 70 indicates illumination that is a light source around the surveillance camera 2, and 2a and 70a indicate an image of the surveillance camera 2 (camera image) and an image of the illumination 70 (illumination image), respectively.

After the above-described three-dimensional position acquisition processing of the monitoring camera 2 is finished, as shown in FIG. is tapped, the SoC 21 of the smartphone 4 acquires the three-dimensional position of the lighting 70 based on the installation information acquisition application 32 and the AR function of the AR application 34 (see FIG. 3). Specifically, in addition to the image 69 captured by the built-in camera 24 of the smartphone 4, the position detected by the touch panel display 26 on the touch panel display 26 tapped by the installation worker and the position of the smartphone 4 captured by the GPS sensor. The three-dimensional position of the lighting 70 is acquired based on the information, the shooting direction of the built-in camera 24 acquired using the electronic compass, the inclination of the smartphone 4 acquired using the acceleration sensor, and the like.

Further, when the acquisition process of the three-dimensional position of the lighting 70 is completed, the SoC 21 of the smartphone 4 superimposes the photographed image 69 on the touch panel display 26 as shown in FIG. 11 based on the installation information acquisition application 32. to display the lighting information input window 71 . The operator who installs the monitoring camera 2 inputs information about the lighting 70 such as the type, intensity, and color by pointing the lighting information input window 71 displayed on the touch panel display 26 with a finger. can do.

Next, with reference to FIGS. 12(a) and 12(b), an example of a camera installation information acquisition method when the surveillance camera 2 is provided with the installation information acquisition unit 51 (see FIG. 5) will be described. When installing or adjusting the monitoring camera 2, the installation worker places a board 72 printed with a two-dimensional barcode 73 shown in FIG. Take a picture with surveillance camera 2. At this time, the installation worker aligns the vertical direction of the two-dimensional barcode 73 with the Z-axis (see FIG. 12(b)) of the floor plan map of the store, and the horizontal direction of the two-dimensional barcode 73 with the X-axis of the floor plan map of the store. Place the camera so that it matches the In this state, the CPU of the surveillance camera 2 (the installation information acquisition unit 51 of the surveillance camera 2) recognizes the two-dimensional barcode 73 in the captured image, thereby determining the installation height position (height from the floor) of the surveillance camera 2 and the installation position. Estimate the angle. The captured image of the board 72 (on which the two-dimensional barcode 73 is printed) by the monitoring camera 2 is displayed on the touch panel display 44 (display unit 64) of the tablet 5. When the dimensional barcode 73 is recognized, as shown in FIG. 12(a), the id given to the recognized two-dimensional barcode 73 (in the example of FIG. 12(a), id=[0, 4, 11, 12]) is displayed, and a detection mark 75 is displayed at the upper right corner of each mark 74 arranged at the top, bottom, left, and right positions of the two-dimensional bar code 73 . The above id=[0, 4, 11, 12] of the two-dimensional barcode 73 indicates that the ids of the top, bottom, left, and right marks 74 in the two-dimensional barcode 73 are 0, 4, 11, and 12, respectively. .

The CPU of the monitoring camera 2 installs the monitoring camera 2 based on how the top, bottom, left, and right marks 74 in the two-dimensional barcode 73 are displayed in the captured image on the touch panel display 44 of the tablet 5. Estimate the height position and installation angle (inclination). More specifically, the CPU (installation information acquisition unit 51) of the surveillance camera 2 determines how the top, bottom, left, and right marks 74 of the two-dimensional barcode 73 in the captured image are displayed (the top, bottom, left, and right marks in the captured image). 74), the installation height position (Y coordinate) and installation angle (inclination) of the monitoring camera 2 are estimated. In addition, the CPU of the monitoring camera 2 (the installation information acquisition unit 51 of the monitoring camera 2), among the installation positions of the monitoring camera 2, the position of the monitoring camera 2 on the XZ coordinate plane (in the store plane map) is displayed on the touch panel display of the tablet 5. A floor plan map of the store is displayed on 44, and the installer of the monitoring camera 2 is prompted to specify and input the position (X coordinate and Z coordinate) of the monitoring camera 2 on the floor map of the store. . In response to this, the installation operator of the monitoring camera 2 inputs the position (X coordinate and Z coordinate) of the monitoring camera 2, so that the CPU of the monitoring camera 2 (the installation information acquisition unit 51 thereof) performs monitoring. The X and Z coordinates of camera 2 can be obtained.

The CPU of the monitoring camera 2 described above can perform camera calibration using the photographed image of the chess board. Note that the CPU 11 of the analysis box 3 may use the image of the chess board captured by the monitoring camera 2 to perform the camera calibration described above.

Next, with reference to FIG. 13, processing instructed from the superimposed image generation search screen 78 displayed on the touch panel display 44 of the tablet 5 will be described. The superimposed image generation search screen 78 is provided with a superimposed image generation button 79, a similar camera image display button 80 of the main store (the same store S), and a similar camera image display button 81 of other stores.

First, superimposed image generation processing performed when the superimposed image generation button 79 is pressed will be described. When the installer of the monitoring camera 2 presses the superimposed image generation button 79, the superimposing unit 58 of the analysis box 3 adds each frame image input from the monitoring camera 2 to each frame image by the image analysis unit 55. A superimposed image is generated by superimposing the information of the analysis result, and the superimposed video composed of the superimposed image is stored in the video DB 54. The camera installation information sent from the acquiring unit 52 is written in the installation information DB 60 in association with the video start time of the superimposed image generated by the superimposing unit 58 . The SoC 41 (see FIG. 4) of the tablet 5 displays the superimposed video recording screen 83 shown in FIG. , and when recording of the superimposed video in the video DB 54 is completed, a superimposed video recording completion screen 84 shown in FIG.

Next, the similar camera image display processing in the same store, which is performed when the similar camera image display button 80 of the main store in FIG. 13 is pressed, will be described. When the installation operator of the monitoring camera 2 presses the similar camera video display button 80 of the main store, the SoC 41 of the tablet 5 displays the store shown in FIG. 16 based on the superimposed video generation confirmation program 47 (see FIG. The inner camera installation state selection screen 85 is displayed. The in-store camera installation state selection screen 85 is a screen in which superimposed images 86 in a plurality of installation states of the monitoring cameras 2 installed in the store where the tablet 5 is placed are displayed side by side. This in-store camera installation state selection screen 85 is displayed on a superimposed image 86 in each installation state of the surveillance camera 2 (installation state 1, installation state 2, etc. in the example of FIG. 16) by the installer of the surveillance camera 2. This is a screen for selecting the image of the monitoring camera 2 that is closest to the installation situation to search for, based on the result of inference (results of human detection and face detection). Each superimposed image 86 displayed on the in-store camera installation state selection screen 85 is obtained by adjusting (changing) the installation state of the same monitoring camera 2, such as the installation position and installation angle, by the installer of the surveillance camera 2. However, in each installation situation, the superimposed image generated when the superimposed image generation button 79 (see FIG. 13) on the superimposed image generation search screen 78 is pressed, or the installation adjustment of another surveillance camera 2 in the same store ( This is a superimposed image generated when the superimposed image generation button 79 is pressed in each installation situation).

When the installer of the monitoring camera 2 presses the playback button 86d below the superimposed video 86 on the in-store camera installation status selection screen 85, the superimposed video 86 is played back. In each frame image forming the superimposed video 86, a person detection (bounding) box 86a indicating detection of a person and a face detection (bounding) box 86b indicating detection of a face are displayed. The human detection (bounding) box 86a and the face detection (bounding) box 86b indicating face detection are the "inference result" described in S9 of FIG. indicate the inference result of human detection and the inference result of face detection, respectively. These inference results (person detection box 86a and face detection box 86b) correspond to "information of analysis results of each frame image" in the claims. 86c in the superimposed image 86 indicates an erroneously detected face (bounding) box corresponding to a face erroneously detected by the image analysis unit 55. FIG.

Below each superimposed video 86 on the in-store camera installation status selection screen 85, camera installation information 87 indicating the installation status of the monitoring camera 2 at the time of generation of each superimposed video 86 and statistical information of each superimposed video 86 are displayed. 88 is displayed. The SoC 41 of the tablet 5 targets the video DB 54, the installation information DB 60, and the analysis result DB 56 of the analysis box 3 installed in the same store S as the tablet 5, and superimposes the video acquired from the video DB 54 and the installation information DB 60. Based on the camera installation information acquired from , and the statistical information acquired from the analysis result DB 56, the superimposed image 86, the camera installation information 87, and the statistical information 88 are displayed.

The camera installation information 87 on the in-store camera installation status selection screen 85 in the example of FIG. The camera installation information on the in-store camera installation status selection screen may include information such as the light source status around the monitoring camera 2 and the detection location of the detection target. Statistical information 88 on the in-store camera installation status selection screen 85 in the example of FIG. However, the statistical information displayed on the in-store camera installation state selection screen 85 is not limited to these. For example, an operator who installs the monitoring camera 2 can select the person detection rate, face detection rate, and upward direction detection rate (from bottom to top) for the corresponding superimposed image 86 from the UI screen on the touch panel display 44 of the tablet 5. , the SoC 41 of the tablet 5 inputs these statistical information to the analysis result DB 56 and displays them as statistical information on the in-store camera installation status selection screen 85. can be Also, when the installation operator of the surveillance camera 2 manually inputs the evaluation value for the corresponding superimposed image 86 from the UI screen on the touch panel display 44 of the tablet 5, the SoC 41 of the tablet 5 receives this evaluation value. The value may be input to the analysis result DB 56 as statistical information and then displayed as statistical information on the in-store camera installation state selection screen 85 .

The installation worker of the surveillance camera 2 selects an installation situation (installation situation 1 in FIG. , installation status 2, etc.), the SoC 41 of the tablet 5 (query issuing unit 67) is installed in the analysis box 3 (query accepting unit 61) installed in the same store S as the tablet 5. A query is issued using the camera installation information corresponding to the installation situation selected by the operator as a key. As a result, the CPU 11 of the analysis box 3 searches for camera installation information similar to the key camera installation information from among the camera installation information stored in the installation information DB 60, and associates the camera installation information with the camera installation information. The superimposed video is read from the video DB 54 using the video start time (generation start time of the superimposed video) as a key, and the read superimposed video is displayed together with the retrieved camera installation information and statistical information corresponding to the superimposed video. , to the tablet 5 side. Then, when the SoC 41 of the tablet 5 receives the superimposed video, the camera installation information, and the statistical information, based on these data, the in-store camera video search result display screen 89 shown in FIG. Displayed on the display 44 .

The in-store camera image search result display screen 89 displays one or more superimposed images 91 with installation conditions close to the installation conditions selected on the in-store camera installation condition selection screen 85, and displays the installation conditions at the time of generation of the superimposed images 91. It is a screen displayed side by side with the camera installation information corresponding to the situation and the statistical information corresponding to the superimposed video. This in-store camera video search result display screen 89 includes the same information as the in-store camera installation status selection screen 85 (superimposed video 91, camera installation information 92, and statistical information 93), as well as the in-store camera installation status. A search key 90 that is camera installation information corresponding to the installation situation selected on the selection screen 85, "search results" (number of results) when a query is issued using this search key 90 as a key, and video identification information 94. to display. The image identification information 94 includes camera identification information (in the example of the figure, the “entrance” (camera of)), the identification number of the superimposed image captured by the monitoring camera 2 corresponding to this camera identification information, and the degree of similarity ( Examples include 100%, 98.5%, and 99%). The installer of the monitoring camera 2 sees the inference results (results of human detection and face detection) and statistical information 93 in each superimposed image 91 displayed on the in-store camera image search result display screen 89, and determines which ( It is possible to judge whether the installation situation of the monitoring camera 2 is favorable. Also, the installer of the surveillance camera 2 shows the inference results and statistical information 93 in each superimposed image 91 to the person in charge of the store S (hereinafter referred to as "customer") where the surveillance camera 2 is installed. It is possible to easily coordinate with the customer how to install (how to set the installation conditions such as the installation position and installation angle of the monitoring camera 2).

Next, the similar camera image display processing of other stores, which is performed when the similar camera image display button 81 of other stores in FIG. 13 is pressed, will be described. When the installer of the surveillance camera 2 presses the similar camera video display button 81 of the other store, the SoC 41 of the tablet 5 displays the other store shown in FIG. 18 based on the superimposed video generation confirmation program 47 (see FIG. A store camera installation state selection screen 95 is displayed. The other store camera installation state selection screen 95 displays superimposed images 86 in a plurality of installation states of the surveillance cameras 2 installed in stores (locations different from the store S) other than the store S where the tablet 5 is placed. This is the screen with This other store camera installation state selection screen 95 allows the installer of the monitoring camera 2 to select the inference result (human detection, face detection, etc.) displayed on the superimposed video 96 in each installation state of the monitoring camera 2 installed in the other store. This is a screen for viewing the result) and selecting the image of the surveillance camera 2 that is closest to the installation situation to search for. In addition, superimposed images 96 displayed on the other store camera installation state selection screen 95 are superimposed images instructed to be generated by installation workers of the monitoring cameras 2 of the other stores when adjusting the installation of the monitoring cameras 2 installed in the other stores. It is a video.

Information (superimposed image 96, camera installation information 97, and statistical information 98) similar to that of the in-store camera installation status selection screen 85 is displayed on the other store camera installation status selection screen 95 described above. 96c in the superimposed image 96 indicates an erroneously detected face (bounding) box corresponding to a face erroneously detected by (the image analysis unit 55 of) the analysis box 3 of another store.

The installation status of the surveillance camera 2 that the installation worker of the surveillance camera 2 wants to adopt is close to the installation status of the surveillance camera 2 that he/she wants to adopt among the plurality of installation situations of the surveillance cameras 2 of other stores displayed on the touch panel display 44 of the tablet 5.例文帳に追加When (installation status 1, installation status 2, etc. in FIG. 18) is tapped and selected, the SoC 41 of the tablet 5 (query issuing unit 67) is an analysis box installed in a store (another store) different from the tablet 5. 3 (the query reception unit 61 of) is issued a query using the camera installation information corresponding to the installation situation selected by the installation worker as a key. As a result, the CPU 11 of the analysis box 3 of the other store searches for camera installation information similar to the key camera installation information from among the camera installation information stored in the installation information DB 60 of the analysis box 3, and The superimposed video linked to the camera installation information is read from the video DB 54 using the video start time as a key, and the read superimposed video is used as the searched camera installation information and the statistical information corresponding to the superimposed video. Together, it is sent to the tablet 5 side. Then, when the SoC 41 of the tablet 5 receives the superimposed video of the other store, the camera installation information, and the statistical information, based on these data, displays the other store camera video search result display screen 99 shown in FIG. 5 is displayed on the touch panel display 44 .

The other store camera image search result display screen 99 displays one or more superimposed images 101 generated in an installation state close to the installation state selected on the other store camera installation state selection screen 95 . This is a screen displayed side by side with camera installation information 103 corresponding to the current installation situation and statistical information 104 corresponding to the superimposed image 101 described above. The other store camera image search result display screen 99 displays the same information as the in-store camera image search result display screen 89 (search key 100, superimposed image 101, image identification information 102, camera installation information 103, and statistical information). 104) is displayed. The operator who installs the monitoring camera 2 sees the inference results (results of human detection and face detection) and statistical information 104 in each superimposed video 101 displayed on the other store camera video search result display screen 99, and sees which ( It is possible to judge whether the installation situation of the monitoring camera 2 is favorable. Therefore, if a certain amount of data is accumulated in the databases (video DB 54, analysis result DB 56, and installation information DB 60) of the analysis box 3 of the other store, the installer of the surveillance camera 2 can use these data (superimposed video, camera Installation information, statistical information), the installation environment of the surveillance camera 2 that you are about to install (installation position of the surveillance camera 2 (height from the floor), light source conditions around the surveillance camera 2, and detection target , etc.), it is possible to search for the installation angle of the camera for achieving high accuracy.

As described above, according to the monitoring camera installation support system 1 of the present embodiment, a superimposed image is generated by superimposing the information of the analysis result of each frame image on each frame image input from the monitoring camera 2. A superimposed video composed of superimposed images is displayed together with camera installation information (for example, information on the installation position and installation angle of the monitoring camera 2) linked to this superimposed video. As a result, when an installation worker of the surveillance camera 2 adjusts (changes) the installation status of the surveillance camera 2 such as the installation position and installation angle of the surveillance camera 2 at the installation site of the surveillance camera 2, the installation status can be changed. Along with the camera installation information shown, the information of the analysis result superimposed on each frame image of the superimposed video (for example, the (inference) result of human detection and face detection) is seen, and the analysis result is obtained by adjusting the installation status of the surveillance camera 2. changes in accuracy can be easily checked.

16 to 19 (in-store camera installation status selection screen 85, in-store camera video search result display screen 89, other store camera installation status selection screen, etc.). As shown in the screen 95 and the other store camera image search result display screen 99), the superimposed images in a plurality of installation situations of the monitoring cameras 2 are displayed side by side. Accordingly, the operator who installs the monitoring camera 2 can compare and confirm the accuracy of the analysis result in a plurality of installation situations of the monitoring camera 2 .

Further, according to the monitoring camera installation support system 1 of the present embodiment, the current The camera installation information corresponding to the installation situation similar to the installation situation of the monitoring camera 2 to be installed is searched, and the superimposed image linked with the searched camera installation information is displayed on the tablet 5 together with the searched camera installation information. The information is displayed on the touch panel display 44. As a result, the installation worker of the monitoring camera 2 can refer to the installation status of the monitoring camera 2 in other stores and the analysis result (inference result) of the superimposed video in the installation status, and 2 installation status (installation position and installation angle) can be determined.

Further, according to the monitoring camera installation support system 1 of the present embodiment, based on information obtained from various sensors (built-in camera 40, GPS sensor, electronic compass, and acceleration sensor) mounted on the smartphone 4, monitoring Information on the installation position and installation angle of the camera 2 can be obtained. As a result, compared to the case where the installation operator of the surveillance camera 2 manually inputs information on the installation position and installation angle of the surveillance camera 2 from the touch panel display 26 of the smartphone 4 (or the touch panel display 44 of the tablet 5), the installation Worker convenience can be improved.

Further, according to the surveillance camera installation support system 1 of the present embodiment, the board 72 (see FIG. 12A) of the two-dimensional bar code 73 installed in the imaging area of the surveillance camera 2 is photographed by the surveillance camera 2. Information on the installation position and installation angle of the monitoring camera 2 is acquired based on the image. As a result, compared to the case where the installation operator of the surveillance camera 2 manually inputs information on the installation position and installation angle of the surveillance camera 2 from the touch panel display 26 of the smartphone 4 (or the touch panel display 44 of the tablet 5), the installation Worker convenience can be improved.

Further, according to the surveillance camera installation support system 1 of the present embodiment, the installation information acquisition unit 51 of the surveillance camera 2 determines the installation height position of the surveillance camera 2 based on the image of each mark 74 of the two-dimensional barcode 73 . and the installation angle can be estimated. As a result, compared to the case where the installation worker of the surveillance camera 2 manually inputs information on the installation height position and installation angle of the surveillance camera 2 from the touch panel display 26 of the smartphone 4 (or the touch panel display 44 of the tablet 5). , the convenience of the installation worker can be improved.

Further, according to the analysis box 3 of the present embodiment, a superimposed image is generated by superimposing the information of the analysis result of each frame image on each frame image input from the monitoring camera 2, and the superimposed image is composed of these superimposed images. The superimposed video is output and displayed on (the touch panel display 44 of) the tablet 5 together with camera installation information (for example, information on the installation position and installation angle of the monitoring camera 2) linked to the superimposed video. As a result, when an installation worker of the surveillance camera 2 adjusts (changes) the installation status of the surveillance camera 2 such as the installation position and installation angle of the surveillance camera 2 at the installation site of the surveillance camera 2, the installation status can be changed. Along with the camera installation information shown, the analysis result information superimposed on each frame image of the superimposed video (for example, the (inference) results of human detection and face detection) is viewed on the touch panel display 44, and the installation status of the monitoring camera 2 is adjusted. changes in accuracy of analysis results can be easily confirmed.

Variant:
The present invention is not limited to the configuration of each of the above embodiments, and various modifications are possible without changing the gist of the invention. Next, modified examples of the present invention will be described.

Variant 1:
In the above embodiment, an example in which the smartphone 4 is provided with the installation information acquisition unit 52 has been mainly described. An installation information acquisition unit may be provided. As a result, the user can use the UI of the tablet to manually input the camera installation information indicating the installation status of the surveillance camera at the time of shooting, or use the methods shown in FIGS. Installation information can be acquired. Note that the monitoring camera installation support system 1 does not need to include the smartphone 4 when the tablet is provided with the installation information acquisition unit.

Modification 2:
In the above-described embodiment, the CPU of the monitoring camera 2 determines how the top, bottom, left, and right marks 74 of the two-dimensional barcode 73 are displayed in the captured image on the touch panel display 44 of the tablet 5. , an example of estimating the installation height position and installation angle of the surveillance camera 2 has been described, but the surveillance camera is equipped with sensors (GPS sensor, electronic compass, acceleration sensor, angular velocity sensor) for detecting the installation position and installation angle. etc.) to estimate the installation position and installation angle of the monitoring camera based on the output values of these sensors.

Variant 3:
Further, in the above embodiment, an example in which the CPU of the surveillance camera 2 estimates the installation height position and the installation angle of the surveillance camera 2 based on the image of each mark 74 of the two-dimensional barcode 73 is shown. . However, the CPU of the surveillance camera may estimate the installation height position and installation angle of the surveillance camera based on the image of the mark other than the two-dimensional bar code.

Variant 4:
In the above embodiment, the analysis box 3 is configured to include the video acquisition unit 53 and the image analysis unit 55. However, the configuration is not limited to this. An AI (Artificial Intelligence) camera having a The function of the unit and the image analysis unit may be provided. Further, the above AI camera may be provided with the functions of the video acquisition unit, the video DB, the image analysis unit, the analysis result DB, and the analysis result superimposition unit that the analysis box has in the above embodiment. good.

Variant 5:
In the above embodiment, an example in which the tablet 5 includes the image confirmation unit 63 has been described. You may make it provide a part.

1 Surveillance camera installation support system 2 Surveillance camera 3 Analysis box (surveillance camera installation support device)
4 Smartphone (mobile terminal)
17 App package (main part of the "surveillance camera installation support program")
24 Built-in camera ("camera" in "mobile terminal")
32 Installation information acquisition application (part of the "surveillance camera installation support program")
47 Superimposed image generation confirmation program (a part of "surveillance camera installation support program")
51 installation information acquisition unit (installation information acquisition means)
52 Installation information acquisition unit (installation information acquisition means)
54 video DB (superimposed video storage means)
55 Image analysis unit (image analysis means)
57 Analysis result superimposition unit (analysis result superimposition means)
60 Installation information DB (installation information storage means)
64 display unit (video display means)
65 camera installation information search unit (camera installation information search means)
68 Installation information acquisition unit (installation information acquisition means)
72 board 73 two-dimensional barcode 74 mark

Claims (10)

In a monitoring camera installation support system for supporting the installation of a monitoring camera for shooting a predetermined shooting area,
installation information acquisition means for acquiring camera installation information indicating the installation status of the surveillance camera at the time of shooting;
image analysis means for analyzing each frame image input from the surveillance camera in the installation situation;
analysis result superimposition means for generating a superimposed image in which information of the analysis result of each frame image by the image analysis means is superimposed on each frame image;
superimposed video storage means for storing a superimposed video composed of superimposed images generated by the analysis result superimposing means in the installation situation;
installation information storage means for storing the camera installation information obtained by the installation information acquisition means at the time of shooting of the surveillance camera in association with the superimposed video stored in the superimposed video storage means;
A monitoring camera installation support system, comprising video display means for displaying the superimposed video stored in the superimposed video storage means together with the camera installation information linked to the superimposed video stored in the installation information storage means. 2. The surveillance camera installation support system according to claim 1, wherein said image display means displays superimposed images in a plurality of installation situations of said surveillance cameras side by side. Similar to the installation situation of the surveillance camera to be installed this time, from among the camera installation information stored in the installation information storage means in another surveillance camera installation assistance system installed at a location different from the surveillance camera installation assistance system further comprising a camera installation information search means for searching for camera installation information corresponding to the installed situation,
3. The image display means displays a superimposed image associated with the camera installation information retrieved by the camera installation information retrieval means together with the retrieved camera installation information. surveillance camera installation support system. 4. The surveillance camera installation support system according to claim 3, wherein the image display means displays, in parallel, each of the superimposed images associated with the plurality of camera installation information retrieved by the camera installation information retrieval means. . 5. The surveillance camera installation support system according to claim 1, wherein the camera installation information includes information on installation positions and installation angles of the surveillance cameras. The surveillance camera installation support system includes a mobile terminal,
6. The apparatus according to claim 5, wherein said installation information acquiring means acquires information on the installation position and installation angle of said surveillance camera based on information obtained from a sensor including a camera mounted on said mobile terminal. Surveillance camera installation support system described. The installation information acquisition means acquires information on the installation position and installation angle of the surveillance camera based on an image of a two-dimensional bar code board installed in the shooting area taken by the surveillance camera. The monitoring camera installation support system according to claim 5. 8. The surveillance camera installation support according to claim 7, wherein the installation information acquiring means estimates the installation height position and installation angle of the surveillance camera based on the image of each mark of the two-dimensional barcode. system. In a monitoring camera installation support device for supporting installation of a monitoring camera that captures images of a predetermined shooting area,
image analysis means for analyzing each frame image input from the surveillance camera;
analysis result superimposition means for generating a superimposed image in which information of the analysis result of each frame image by the image analysis means is superimposed on each frame image;
superimposed video storage means for storing a superimposed video formed from each superimposed image generated by the analysis result superimposing means;
installation information storage means for storing camera installation information at the time of shooting of the surveillance camera in association with the superimposed video stored in the superimposed video storage means;
a video output means for outputting the superimposed video stored in the superimposed video storage means to display means together with the camera installation information linked to the superimposed video stored in the installation information storage means. support equipment. A monitoring camera installation support program for supporting installation of a monitoring camera for shooting a predetermined shooting area,
the computer,
installation information acquisition means for acquiring camera installation information indicating the installation status of the surveillance camera at the time of shooting;
image analysis means for analyzing each frame image input from the surveillance camera in the installation situation;
analysis result superimposition means for generating a superimposed image in which information of the analysis result of each frame image by the image analysis means is superimposed on each frame image;
superimposed video storage means for storing a superimposed video composed of superimposed images generated by the analysis result superimposing means in the installation situation;
installation information storage means for storing the camera installation information obtained by the installation information acquisition means at the time of shooting of the surveillance camera in association with the superimposed video stored in the superimposed video storage means;
A surveillance camera installation support program for functioning as video display means for displaying the superimposed video stored in the superimposed video storage means together with the camera installation information linked to the superimposed video stored in the installation information storage means. .

Images