Surveillance support system, method, and program

Abstract

To support a surveillant in efficiently detecting a suspicious person or the like by appropriately displaying images to be focused on from surveillance images.SOLUTION: A camera unit (100) includes: a detection unit (151) that detects a specific action of a person in layer images by a detection method corresponding to respective layers, for a plurality of layer images corresponding to a plurality of layers that are captured by a camera at a predetermined time point and that differ in distance from the camera; an identification unit (152) that, when the specific action is detected, identifies an image group captured by the camera during a period from a first time before the predetermined time point to the predetermined time point; and a setting unit (153) that sets, for the identified image group, a priority for transmission that is higher than a normal priority of images.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a monitoring support system, method, and program. [Background technology]

[0002] Generally, surveillance cameras sample some frame images from the captured video data. The monitor device then receives the frame image and transmits it to the monitor device. The image is displayed on the screen at intervals longer than the frame rate set on the monitor screen. To the observer, the video footage captured by the surveillance camera appears to be played back frame by frame. This may result in a delay in the supervisor's detection of suspicious individuals.

[0003] In Patent Document 1, when the number of people in the captured image is equal to or greater than a predetermined value, the camera captures images from nearby cameras. Techniques for modifying shadow regions are disclosed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2005-117542 Summary of the Invention [Problem to be solved by the invention]

[0005] As mentioned above, the monitor of the video taken by the surveillance camera installed in the store or the street There is a problem that it is difficult for the monitor to detect suspicious individuals from the images displayed on the device. It should be noted that the technology disclosed in the above-mentioned Patent Document 1 does not detect suspicious movements.

[0006] The present disclosure has been made in view of the above-mentioned problems, and provides a method for appropriately extracting noteworthy images from surveillance images. By clearly displaying the information, the monitor can efficiently detect suspicious individuals. The object is to provide a support system, method and program. [Means for solving the problem]

[0007] The first aspect of the present disclosure is a method for detecting an object captured by a camera at a predetermined time and a distance from the camera. For multiple layer images corresponding to multiple layers with different a detection unit that detects a specific behavior of a person in the layer image by If so, images taken by the camera between the predetermined time and a first time A specifying unit for specifying a group and a priority for transmitting the group with priority over images with normal priority are provided. and a setting unit that sets the specified image group.

[0008] A second aspect of the present disclosure is a computer that acquires a photograph taken by a camera at a predetermined time and For multiple layer images corresponding to multiple layers with different distances from the camera, detecting a specific behavior of a person in the layer image by a detection method according to the When the specific behavior is detected, the camera and a step of identifying a group of images taken by the camera and transmitting the images with priority over images with normal priority. and setting a priority for the identified group of images to be monitored. To provide.

[0009] A third aspect of the present disclosure is a method for detecting an object captured by a camera at a predetermined time and a distance from the camera. For multiple layer images corresponding to multiple layers with different detecting a specific behavior of a person in the layer image by When the image is displayed, images taken by the camera from the predetermined time to a first time A step of identifying an image group and assigning a priority to the image group so that the image group is transmitted with priority over normal priority images. a step of setting the specified image group; and a monitoring support program for causing a computer to execute the step. Provide grams. [Effects of the Invention]

[0010] This disclosure allows a monitor to detect suspicious activity by appropriately displaying noteworthy images from surveillance footage. Surveillance support system, method and program for supporting efficient detection of suspects, etc. can be provided. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a block diagram showing an overall configuration including a monitoring support system according to a first embodiment. [Figure 2] FIG. 1 is a block diagram showing the configuration of a camera unit according to the first embodiment. [Figure 3] FIG. 2 is a block diagram showing the configuration of a switch unit according to the first embodiment. [Figure 4] FIG. 2 is a block diagram showing the configuration of a monitor unit according to the first embodiment. [Figure 5] 5 is a flowchart showing the flow of a monitoring support process in the camera unit according to the first embodiment. [Figure 6] 5 is a flowchart showing the flow of a monitoring support process in the camera unit according to the first embodiment. [Figure 7] 10A and 10B are diagrams for explaining the concept of pattern matching and layer images according to the distance from the camera according to the first embodiment. [Figure 8]FIG. 10 is a diagram showing an example of a frame image transmitted with normal priority and displayed on a monitor unit according to the first embodiment. [Figure 9] FIG. 10 is a diagram showing an example of a frame image transmitted when the upper region matches a pattern (higher priority than normal) and displayed on a monitor unit according to the first embodiment. [Figure 10] FIG. 10 is a diagram showing an example of a frame image transmitted when the lower area matches a pattern (higher priority than normal) and displayed on a monitor unit according to the first embodiment. [Figure 11] FIG. 10 is a diagram showing an example of a frame image transmitted (with a higher priority than normal) and displayed on a monitor unit when a person moves from an upper region to a lower region according to the first embodiment. [Figure 12] FIG. 10 is a block diagram showing a hardware configuration of a monitoring support device according to another embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] Specific embodiments of the present disclosure will be described in detail below with reference to the drawings. In each drawing, the same elements are given the same reference numerals, and for clarity of explanation, Duplicate explanations will be omitted accordingly.

[0013] <Embodiment 1> FIG. 1 is a block diagram showing the overall configuration including a monitoring support system 1000 according to the first embodiment. The monitoring support system 1000 includes camera units 100-1 to 100-n (n is a natural number equal to or greater than 1.) and a switch unit 200. -1 to 300-m (m is a natural number of 1 or more) are images captured by the surveillance support system 1000. The monitor then displays the captured video data on the screen as appropriate. The monitoring support system 100 monitors the screen of the person in question and detects a suspicious person or the like on the screen. 0 allows the monitor to efficiently detect suspicious individuals by appropriately displaying noteworthy images from the surveillance video. The camera unit 100 is an information system for assisting in the efficient detection of Each of the monitor units 100-1 to 100-n is connected to the switch unit 200. Each of the ports 300 - 1 to 300 - m is connected to the switch unit 200 .

[0014] The camera unit 100-1 and the like are surveillance support devices including surveillance cameras installed in stores and on the street. Each of the camera units 100-1 to 100-n has the same configuration. Therefore, in the following description, when there is no need to distinguish between them, we will simply refer to them as camera units. 100. In addition, each of the monitor units 300-1 to 300-m are equivalent in construction, and in the following description, when there is no need to distinguish between them, hereinafter referred to as monitor unit 300.

[0015] The camera unit 100 takes pictures at predetermined intervals, sets priorities for the pictures taken, and switches between them. In particular, the camera unit 100 transmits the captured image to the camera unit 200. The system captures multiple layer images according to the distance from the object, and detects specific human behavior from each layer image. If this is the case, the priority of the frame image corresponding to the layer image is set higher than normal.

[0016] FIG. 2 is a block diagram showing the configuration of the camera unit 100 according to the first embodiment. The camera unit 100 can be considered an example of a surveillance support system or surveillance support device. The unit 100 includes a camera 110, a depth sensor 120, a timer 130, and an image buffer 1 40, a detection unit 151, a determination unit 152, a setting unit 153, and a transmission unit 154. 10 performs photography at a predetermined interval, for example, at a frame rate, and acquires the photographed frame images. The depth sensor 120 is a sensor that measures the distance between the camera 110 and the subject as depth. Therefore, the camera 110 and the depth sensor 120 are used together as a TOF (Time Of Flight) sensor. At the very least, the camera 110 may be a camera that captures frames captured at a given time. Along with the image, the camera 110 and the subject are detected based on the depth measured by the depth sensor 120. Here, the layer images are acquired at the same time. It is an image of a specific area within a captured frame image. However, it is not possible to use multiple images captured at the same time. The layer images may overlap each other in the frame image. The acquired frame image and the corresponding layer images are stored in the image buffer 140. do.

[0017] The image buffer 140 stores frame images 141, 142, For example, the frame image 141 is a storage area for storing the layer image 1411. , 1412, etc. Alternatively, the frame image 141 includes layer images 1411, 141 2... are stored in association with each other. Frame image 142 and subsequent images have the same structure. The image buffer 140 is, for example, a semiconductor memory.

[0018] The timer 130 periodically outputs a sampling pulse to the detection unit 151. The interval between sampling pulses is set to be sufficiently longer than the interval between images taken by the camera 110. For example, if the camera 110 is shooting at 60 fps (frames per second), The ring pulse is output at intervals of three frames or more. For example, it may be an analog circuit or a digital circuit.

[0019] The detection unit 151 receives a sampling pulse from the timer 130. For each layer, the detection unit 151 performs processing such as image recognition on the layer image. The layer-specific detection method detects specific actions of people in the layer image of the corresponding layer. For example, the detection unit 151 detects whether a person in the layer image is likely to be a suspicious person. If it is determined that there is a possibility, the specific behavior is detected. 51 determines whether a person in a layer image is engaging in suspicious behavior or actions, and When it is determined that the user is performing a certain action, the specific action is detected.

[0020] The detection unit 151 also uses pattern matching between images to identify people. For example, the detection unit 151 may detect the behavior of the user based on the layer image captured at the first time. The image is compared with a layer image of the same layer taken at a second time later than the first time. Then, the detection unit 151 calculates the difference in data size and the amount of change in pixels. If the difference amount or the like is equal to or greater than a predetermined value, it is determined that the person matches a specific behavior pattern. A specific action may be detected as a person moving. For example, the detection unit 151 detects that the first and second If the positions of people in the layer images at the time are separated by a predetermined distance or more, The person moves between the first time and the second time, and this may be detected as a specific behavior. In addition, the detection unit 151 detects a specific part of a person in the image between the layer images at the first and second times. If the distance or angle changes at a predetermined rate, the person can be identified from the first time to the second time. A specific part may be moved and detected as a specific behavior. Examples include, but are not limited to, the arm, fingertip, and face direction.

[0021] Furthermore, if the layer indicates a distance less than a predetermined distance from the camera 110, the detection unit 151 ,By using pattern matching of specific parts of a person in the ,layer image as a detection method, The detection unit 151 may detect a specific behavior. If the distance is greater than 100m, the pattern matching of the overall position of the person in the layer image is used as the detection method. A specific behavior may be detected by performing the following.

[0022] The detection unit 151 also detects a second time period in each layer according to the distance from the camera 110. The pixels in a predetermined number of layer images corresponding to the layer and taken at different times are Then, the detection unit 151 calculates the average value of the amount of change in the rays. For layer images taken at a predetermined time corresponding to the layer, a detection method corresponding to the layer is used. It is preferable to detect a specific behavior by the above-mentioned first time and second time. The interval is equal to or longer than the interval between shots taken by the camera 110 and within the interval between sampling pulses. The first time and the second time may be different times. For example, the second time may be the time when a frame image including one layer image is captured and the time when the other layer image is captured. The interval between the time when the frame image containing the layer image was captured and the time when the frame image containing the layer image was captured is 100%. The time may be different for each layer.

[0023] When a specific behavior is detected, the identification unit 152 Identify a group of images captured by the camera 110. Here, for example, the image capture time is determined to be a predetermined interval. If the shooting interval is t1, t2, t3, t4, etc., the sampling pulse The timing is assumed to be times t1, t4, etc. In this case, the first The time before is the time when the detection unit 151 received the sampling pulse from the timer 130 last time. The timing of receiving the data, for example, the timing of the most recent reception after time t1, for example, time t4 In other words, the period from the predetermined time point to the first time point is the time t1, t2, or t3. For example, if the shooting interval of the camera 110 is 6 In the case of 0 fps, the period from the predetermined time point to the first time point is, for example, 1 / 20 seconds. This is not limited to this.

[0024] The setting unit 153 sets a priority for transmitting an image with priority over an image with normal priority. The normal priority images are set to the group of images switched from the camera unit 100. The priority level for transmitting to the monitor unit 300 via the switch unit 200 is normal. For example, a part of the frame images of the captured video data is sampled. The frame images sampled at the intervals of 1 / 2000 correspond to normal priority images. Priority is set to send images with priority over priority images, even if no specific behavior is detected. The frame images that have a higher transmission priority than the frame images that have been sent, that is, the frame images that have a higher transmission priority than the frame images that have been sent, Here, the priority may be expressed as a numerical value or a level value. The priority may be the suitability of pattern matching, the importance of suspicious behavior, or the like.

[0025] The transmission unit 154 transmits the image group for which the priority has been set to the display device. The transmitter 154 sends the switch packet to the monitor unit 300 regardless of the priority. The transmitting unit 154 transmits packets of each image to the switch unit 200. In this case, packets with lower priority may be sent. Good too.

[0026] The detecting unit 151, the identifying unit 152, the setting unit 153, and the transmitting unit 154 are functional blocks. Yes, by hardware or software or a combination of hardware and software This is realized.

[0027] FIG. 3 is a block diagram showing the configuration of the switch unit 200 according to the first embodiment. The switch unit 200 includes a buffer 210, a receiver 220, a scheduler 230, and The buffer 210 includes a transmitting unit 240. The buffer 210 receives a plurality of packets from the switch unit 200. The buffer 210 is a storage area for storing packets 211, 212, etc. The packet 211 includes a source address, a destination address, and a In addition to the packet size, the packet 211 also includes a priority 2111. The packet 211 also includes a payload Packet 212 contains compressed data corresponding to a frame image. The payload contains compressed data corresponding to a frame image.

[0028] The receiving unit 220 receives each of the plurality of camera units 100-1 to 100-n. The packet is received via the network interface corresponding to the camera unit. The receiving unit 220 stores the received packets in the buffer 210. The circuit includes a communication interface module with the host.

[0029] The scheduler 230 transmits each packet stored in the buffer 210 according to its priority. Specifically, the scheduler 230 prioritizes packets with high priority. The scheduler 230 is, for example, a hardware circuit. or a processor that executes a scheduler program.

[0030] The transmitter 240 transmits the packets selected by the scheduler 230 to the monitor unit 30. The transmitter 240 transmits the signal to the monitor unit 300 via a communication interface 301. A circuit including a module.

[0031] FIG. 4 is a block diagram showing the configuration of the monitor unit 300 according to the first embodiment. The monitor unit 300 includes a receiver 310, a decoder 320, and a monitor 330. The receiving unit 310 receives a packet from the switch unit 200 via a network interface. The receiving unit 310 receives the bit and outputs it to the decoder 320. The receiving unit 310 includes, for example, a switch unit 200 and a circuit including a communication interface module.

[0032] The decoder 320 receives the compressed data included in the payload of the packet received from the receiver 310. The decoder 320 decodes the data and restores it to image data. The monitor 330 displays the restored image data. The monitor 330 displays the data on a screen. The monitor 330 is, for example, a liquid crystal display device.

[0033] 5 and 6 show the flow of the monitoring support process in the camera unit according to the first embodiment. First, the camera 110 detects the location of the object measured by the depth sensor 120. Based on the depth, multiple layer images taken at a given time are acquired. The image processing unit 10 stores the acquired layer images in the image buffer 140 (S101).

[0034] FIG. 7 shows the layer image and pattern matching according to the distance from the camera according to the first embodiment. In this example, a layer image 41 is a 4 shows a partial area of ​​the frame image, and layer image 42 shows an upper area of ​​the same frame image.

[0035] The layer image 41 is image data in the layer 1 that is relatively close to the camera 110. Layer 1 shows the area less than a predetermined distance from the camera 110. The pixel information indicating that a person is present at position P2 in the image 1 is included. Even if a person in the same level moves his / her arm from position P2-1 to P2-4, Therefore, the layer image 41 shows the position P P2-1 and P2-4 contain pixel information indicating the presence of an arm.

[0036] The layer image 42 is image data in layer 2, which is relatively far from the camera 110. Layer 2 indicates a predetermined distance or more from the camera 110. The layer image 42 shows the Pixel information indicating the presence of a person at position P1-1 or P1-4 in Y2 is included. Also, even if a person at position P1-1 moves to position P1-4, Therefore, the layer image 42 shows the position in layer 2. The pixel information indicating that a person is present at positions P1-1 and P1-4 is included.

[0037] Returning to FIG. 5, the description will be continued. After step S101, for example, the timer 130 starts sampling. Assume that the detecting unit 151 outputs a plurality of level pulses to the detecting unit 151. Then, the detection unit 151 selects an unselected layer from the selected layer (S102). The layer images corresponding to the layer from the most recent to the predetermined time ago are read from the image buffer 140. The detection unit 151 detects the most recent and the previous layer images, that is, the total The total number of layers is two or more, and at least the untransmitted layer images are read. 51 excludes layer images that have already been transmitted. The image data captured before the frame image transmitted may be deleted.

[0038] Then, the detection unit 151 calculates the pixel change amount between successive layer images (S10 4) Here, consecutive layer images are layer images captured at adjacent times. For example, The detection unit 151 encodes the layered image and calculates the bit rate for each picture type. The picture type is the encoding type, such as I picture, B picture, or P picture. Then, the detection unit 151 detects the bit rate for each picture type for the consecutive layer images. Alternatively, the detection unit 151 calculates the pixel amount of each layer image and calculates the difference between the pixel amounts of the connected images. Alternatively, the detection unit 151 calculates the difference in pixel amount between the adjacent layer images. Calculate the data size of the differential frame of the image and calculate the difference in data size after encoding. The pixel change amount is the amount of change in pixel brightness between layer images that exceeds a predetermined value. is the number of pixels.

[0039] Then, the detection unit 151 calculates the average value of the pixel change amounts (S105). For example, If three or more layer images are read in step S103, a pair of adjacent layer images is Then, the detection unit 151 detects two or more pixel changes in step S104. Therefore, the detection unit 151 calculates the amount and difference value of the pixel calculated in step S104. The average value of the cell change amount etc. is calculated. In step S103, two layer images are read out. If so, since there is only one pair of adjacent layer images, step S105 is omitted and step The calculated value in step S104 may be interpreted as an average value.

[0040] Then, the detection unit 151 determines whether the average value is greater than a threshold value corresponding to the layer. (S106). Here, the threshold value corresponding to the layer is, for example, in the case of layer 1, is the pixel change value that may indicate suspicious behavior, and in the case of Layer 2, The amount of movement is the pixel change value that may indicate suspicious behavior. The threshold value can be arbitrarily changed for each layer.

[0041] If it is determined in step S106 that the average value is greater than the threshold value corresponding to the layer, the detection unit 151 selects a detection method corresponding to the selected layer (S107). For example, For Layer 1, the formula is a pattern matching method to detect the movement of a person's arms, In the case of layer 2, a pattern matching method may be used to detect the movement of a person.

[0042] Then, the detection unit 151 detects the most recently captured layer image and the layer image captured immediately before. The image is subjected to pattern matching using the selected detection method (S108). In the case of 1, the detection unit 151 detects the area of ​​the person's arm from each layer image by image recognition, Calculate the difference in the position, length, angle, etc. of the arm area between the two layer images. In this case, the detection unit 151 detects the area of ​​the overall position of the person from each layer image by image recognition. Then, the difference in the detected area between the two layer images is calculated.

[0043] Then, the detection unit 151 determines whether or not the pattern matches (S109). Specifically, the detection unit 151 determines whether the difference calculated in step S108 is equal to or greater than a predetermined value. If the difference is equal to or greater than a predetermined value, the detection unit 151 determines that the pattern matches. For example, in the case of Layer 1, if the difference in the position, length, angle, etc. of the arm area is greater than a predetermined value, The movement of the person is suspicious and matches the pattern. If the difference in position, i.e., the movement distance, is greater than a predetermined value, it is considered suspicious because it is a sudden movement, and the pattern However, examples of pattern matching are not limited to this.

[0044] If it is determined in step S109 that the pattern matches, the detection unit 151 Determines the higher priority than normal for the layer image corresponding to the selected layer in (S110). The frame image 51 is stored in a memory or the like in association with the layer image and the priority. You may record it.

[0045] After step S110 or if it is determined in step S109 that the pattern does not match In this case, the detection unit 151 determines whether all layers have been selected (S111). If it is determined in step S106 that the average value is equal to or less than the threshold value corresponding to the layer, step S111 is also performed. Execute.

[0046] In step S111, not all layers have been selected, that is, there are unselected layers. If so, the process returns to step S102 and the subsequent processes are repeated. If it has been selected, the specifying unit 152 assigns a higher priority than usual to the processed layer image. Specifically, the specifying unit 152 determines whether the degree has been determined (S112). and determines whether the priority associated with the layer image is higher than normal.

[0047] If it is determined in step S112 that a higher priority than normal has been determined, the specifying unit 152 Then, a group of frame images from the predetermined time point to the first time point is identified (S113). is the time when the detection unit 151 receives a sampling pulse after step S101, that is, That is, the time when the frame image or layer image to be processed was captured. 2 is a frame image at a predetermined time point from the image buffer 140 and a first time point from the predetermined time point. The frame images are read as a group of one or more frame images taken up to the point in time between the current time and the current frame image. Therefore, a group of frame images having intervals shorter than the sampling interval is identified.

[0048] Then, the setting unit 153 sets the frame image group packet in step S110. In other words, the setting unit 153 sets the priority determined in the specified frame (S114). The compressed data of each image in the image group is set in the payload of the packet to be sent, and the header The setting unit 153 sets a higher priority than normal to each of the packets. You can do that.

[0049] If it is determined in step S112 that a priority higher than normal has not been determined, the specifying unit 1 The identification unit 152 identifies a frame image at a predetermined time point (S115). 2 reads out a frame image at a predetermined time from the image buffer 140. The unit 153 sets a normal priority to each packet of the identified frame image (S116 That is, the setting unit 153 transmits compressed data obtained by compressing the image of the specified frame image. The setter 15 sets the payload of the trusted packet and sets the normal priority in the header. 3. An image may be divided into multiple packets.

[0050] After step S114 or S116, the transmitter 154 sends each packet to the switch unit At this time, the transmitting unit 154 transmits each packet to the camera 200 in the order of the time of shooting. It may be sent in order.

[0051] The processes from step S104 to step S106 may be omitted. After step S103, step S107 and subsequent steps are executed. This process may be omitted. In this case, the average value is calculated by subtracting the threshold value corresponding to the layer in step S106. If it is determined that the value is larger than the predetermined value, step S110 and subsequent steps are executed.

[0052] Thereafter, the switch unit 200 switches the camera units 100-1 to 100-n Packets received from each are stored in the buffer 210 and sorted according to the priority of each packet. The packets are sent to monitor units 300-1 to 300-m in that order.

[0053] The monitor unit 300 displays image data corresponding to each received packet. Therefore, the video is displayed according to the reception interval. If the packet is sent to the monitor unit 300 according to normal priority, the monitor unit The monitor unit 300 reproduces frame images at sampling intervals. If the interval between receiving packets from the switch unit 200 is long, the network unit 300 The image will be displayed frame by frame for the user.

[0054] On the other hand, the camera unit 100 detects a frame in which a specific behavior is detected by the detection unit 151. The image and a certain number of previous frame images are set to high priority and switched to Therefore, the switch unit 200 transmits the packet with a higher priority than the packet with a normal priority. The packets with higher priority than the others are sent to the monitor unit 300. The monitor unit 300 receives packets with higher priority at shorter intervals than usual. The monitor unit 300 receives the frame image and displays the decoded frame image on the screen. may display video in which suspicious activity is detected at a higher frame rate than normal. Therefore, the observer can observe the movements of a suspected person in more detail and smoothly. Suspicious people can be detected efficiently by visually checking the video.

[0055] FIG. 8 shows the frame image transmitted with normal priority according to the first embodiment on the monitor unit. 300. Here, frame images 51, 53, and 54 are displayed. The upper and lower regions 511 and 512, 531 and 532, respectively, The upper region 511 and the lower region 512 include an upper region 541 and a lower region 542. There may be some overlapping areas. The same applies to other areas. The camera 110 takes photographs at intervals of t1, t2, t3, t4, etc. The sampling interval of the camera unit 100 is assumed to be time t1 followed by time t4. The same applies to other explanations below.

[0056] At time t1, the monitor unit 300 displays a frame image 51. In the upper region 511 of the image 51, a person is present at a position P1-1 far from the camera 110, and in the lower region 512 of the image 51, a person is present at a position P1-2 far from the camera 110. In the area 512, a person is present at a position P2 close to the camera 110, and the person's arm is at a position It is assumed to be present in P2-1.

[0057] After that, the frame image taken at time t4, when the sampling interval has elapsed since time t1, is For 54, the camera unit 100 executes the processes shown in FIGS. At this time, the camera unit 100 compares the frame image 54 with the frame image captured at the most recent time t3. The layer images are compared with the group image 53 to determine whether the patterns match. In this example, the camera unit 100 is located at a position P1-b in the upper region 541 and a position P2-b in the upper region 53 Since the pixel change amount between position P1-a and position P1-b is less than the threshold, The camera unit 100 determines that the amount of movement of the person is small. Position P2 and position P2-1, and pixel change between position P2 and position P2-1 in the lower region 532 Since the amount is below the threshold, it is determined that the arm movement of the person in the lower area, which is layer 1, is small. Therefore, the camera unit 100 sets a normal priority to the frame image 54 and switches it. Conversely, the camera unit 100 transmits the image data to the touch unit 200 at time t3. The switch unit 200 does not transmit the frame image 53. The packet of image 54 is sent to the monitor unit 300. Therefore, the monitor unit 300 At time t4, the frame image 54 is displayed. That is, at time t1, the frame image 51 is displayed. After that, the frame image 51 is not updated at time t2 and t3, and the frame image Therefore, the observer can see the frame images 51 to 54 frame by frame. However, the person in the upper area 541 moves from position P1-1 to position P1-b. Since the user has simply moved to a different location, there is little need to alert the user to this suspicious behavior, so there is no impact on monitoring. is few.

[0058] FIG. 9 shows the case where the upper region of the first embodiment matches the pattern (higher priority than usual). FIG. 10 is a diagram showing an example of a frame image transmitted previously being displayed on the monitor unit 300. The frame image 61 displayed at time t1 is the same as the frame image 51 described above. do.

[0059] After that, at time t4, when the sampling interval has elapsed since time t1, the camera unit 100 is a position P1-4 of the upper region 641 of the frame image 64 and a position P2 of the upper region 631 Since the pixel change amount between P1 and P3 is greater than the threshold, it is determined that the movement amount of the person in the upper area is large. Therefore, the pattern is matched in layer 2, and the camera unit 100 starts frame images 62 to 64 from the time t1 to the time t2, which is the first time before, are identified, and Each packet is sent to the switch unit 200 with a higher priority than normal. The camera unit 100 transmits frame images 62 and 63, which are not subject to transmission at normal priority. 3. Then, the switch unit 200 transmits the frame images 62, 63 and 64. The packet is sent to the monitor unit 300 with priority. After displaying the frame image 51 at time t1, the frame image 62 is displayed at time t2, and the frame image 63 is displayed at time t3. At time t1, frame image 63 is displayed, and at time t2, frame image 54 is displayed. From the top of the image 61 to 64, the people in the upper area are at positions P1-1, P1-2, P1-3, and P1 The movement from -4 to -4 can be seen as a smooth movement. Therefore, the person in the upper area is suspicious. It can efficiently determine whether a person is a

[0060] FIG. 10 shows the case where the lower region of the first embodiment matches the pattern (higher than normal). 10 is a diagram showing an example in which a frame image transmitted via a network (high priority) is displayed on the monitor unit 300. The frame image 71 displayed at time t1 is the same as the frame image 51 described above. Let's say.

[0061] After that, at time t4, when the sampling interval has elapsed since time t1, the camera unit 100 is a position P2-4 of the lower region 742 of the frame image 74 and a position P3 of the lower region 732 The pixel change from P2-3 is greater than the threshold, so the person in the lower area has a lot of arm movement. Therefore, the pattern is matched in layer 1, and the camera unit 100 determines that the frame images 72 to 74 from time t4 to time t2, which is the first time before, are identified, and each packet Each packet is set to a higher priority than normal and sent to the switch unit 200. That is, the camera unit 100 transmits the frame images 72 and 73 that are not subject to transmission at normal priority. The switch unit 200 also transmits the frame images 72, 73 and 74. The packet No. 4 is sent to the monitor unit 300 with priority. 0, after displaying the frame image 71 at time t1, the frame image 72 at time t2, and the frame image 73 at time t At time t3, a frame image 73 is displayed, and at time t4, a frame image 74 is displayed. From frame images 71 to 74, the arms of the person in the lower area are at positions P2-1, P2-2, and P2- 3. The movement of the object to P2-4 can be seen as a smooth movement. In this example, the observer can efficiently determine whether the person is suspicious or not. Between t2 and t3, the person in the upper region moves from position P1-1 to P1-3. This can also be seen.

[0062] FIG. 11 shows the case where a person moves from the upper area to the lower area according to the first embodiment (compared to the usual case). 10 shows an example in which a frame image transmitted (with a higher priority) is displayed on the monitor unit 300. Figure.

[0063] At time t1, the monitor unit 300 displays a frame image 81. In the upper region 811 of the image 81, a person is present at a position P3-1 far from the camera 110, and in the lower region 812, a person is present at a position P3-2 far from the camera 110. It is assumed that no person exists in the area 812.

[0064] After that, at time t4, when the sampling interval has elapsed since time t1, the camera unit 100 is a position P3-4 of the lower region 842 of the frame image 84 and a position P3-5 of the lower region 832 Since the pixel change amount between P3-3 is greater than the threshold, it is determined that the movement amount of the person in the lower area is large. Therefore, the pattern is matched in layer 1, and the camera unit 100 starts frame images 82 to 84 from the time t1 to the time t2, which is the first time before, are identified, and Each packet is sent to the switch unit 200 with a higher priority than normal. The camera unit 100 transmits frame images 82 and 83, which are not subject to transmission at normal priority. 3. Then, the switch unit 200 transmits the frame images 82, 83 and 84. The packet is sent to the monitor unit 300 with priority. After displaying the frame image 81 at time t1, the frame image 82 is displayed at time t2, and the frame image 83 is displayed at time t3. At time t1, a frame image 83 is displayed, and at time t2, a frame image 84 is displayed. From the top of the image 81 to 84, the person in the upper area is at the positions P3-1, P3-2, P3-3, P The movement from 3 to 4 can be seen with smooth movement. Therefore, the person on the screen is a suspicious person. It is possible to efficiently determine whether

[0065] Here, in order to narrow down the target images for surveillance and detect suspicious individuals, It is also possible to train an image recognition algorithm to learn the behavior of examiners. In contrast, in this embodiment, an advanced image recognition algorithm is used. Without the need to train algorithms, the system can identify noteworthy images from surveillance footage and prioritize communications. Therefore, if the possibility of suspicious behavior is low, the communication load and While keeping costs down, if there is a high possibility of suspicious activity, the camera will be displayed at an appropriate frame rate. The image can be displayed.

[0066] <Other embodiments> The above-mentioned camera unit 100 is realized by an information processing device which is a monitoring support device. 12 is a block diagram showing the hardware configuration of a monitoring support device 100a according to another embodiment. 1 is a block diagram showing the configuration of a monitoring support device 100a. 0, timer 130, image buffer 140, control unit 150, storage unit 160, memory 170, and and a communication unit 180. The camera 110, the depth sensor 120, the timer 130, and the image buffer The fan 140 is the same as that shown in FIG. 2 and so a description thereof will be omitted.

[0067] The storage unit 160 is an example of a storage device such as a hard disk or a flash memory. The unit 160 stores a monitoring support program 161. The monitoring support program 161 is The present invention is a computer program that implements the monitoring support process and the like according to the embodiment.

[0068] The memory 170 is a volatile storage device such as a RAM (Random Access Memory), and the control unit The communication unit 180 is a storage area for temporarily storing information during the operation of the monitoring unit 150. The internal configuration of the support device 100a and the communication interface circuit with the switch unit 200 be.

[0069] The control unit 150 is a processor that controls each component of the monitoring support device 100a, i.e., a control device. The control unit 150 transfers the monitoring assistance program 161 from the storage unit 160 to the memory 170. The control unit 150 then reads the monitoring support program 161 and executes the monitoring support program 161. The function of the detecting unit 151, the identifying unit 152, the setting unit 153, the transmitting unit 154, etc. is realized.

[0070] Alternatively, each component of the monitoring support device 100a may be realized by dedicated hardware. In addition, some or all of the components of each device may be general-purpose or dedicated circuits. These may be realized by a circuitry, a processor, or a combination thereof. It may be configured by a single chip or by multiple chips connected via a bus. A part or all of the components of each device may be configured by programming the circuits and the like described above. The processor may be a CPU (Central Processing Unit) or a combination of the CPU and the processor. Processing Unit), GPU (Graphics Processing Unit), FPGA (Field-Program Mabl e Gate Array), quantum processor (quantum computer control chip), etc. can be used. do.

[0071] The present invention has been described above based on the above embodiment, but the present invention does not depend on the configuration of the above embodiment. The present invention is not limited to the above, but is within the scope of the claims of the present application. Of course, this includes various modifications, alterations, and combinations that can be made.

[0072] In the above embodiment, the configuration is explained as a hardware configuration, but the present invention is not limited to this. This disclosure does not mean that any process can be performed by causing a CPU to execute a computer program. This can also be achieved by

[0073] In the above example, when the program is loaded into a computer, the program A set of instructions (or software) that causes a computer to perform one or more of the specified functions. The program may be stored on a non-transitory computer-readable medium or tangible storage. By way of example and not limitation, the present invention may be stored on a computer-readable medium or tangible medium. Storage media include random-access memory (RAM), read-only memory (ROM), and flash memory. solid-state drive (SSD) or other memory technology, CD-ROM, digital versatile disc sc (DVD), Blu-ray (registered trademark) disc or other optical disc storage, magnetic cassette magnetic tape, magnetic disk storage or other magnetic storage devices The program may be transmitted on a temporary computer-readable medium or a communication medium. By way of example, and not limitation, transitory computer-readable media or communication media may include electrical, optical, Includes acoustic or other forms of propagated signals. [Explanation of symbols]

[0074] 1000 Monitoring Support System 100 camera unit 110 Camera 120 Depth Sensor 130 Timer 140 Image Buffer 141 frame images 1411 Layer Images 1412 Layer Images 142 frame images 151 Detector 152 Specific part 153 Setting section 154 Transmitter 200 Switch Unit 210 buffers 211 packets 2111 Priority 212 packets 2121 Priority 220 Receiving unit 230 Scheduler 240 Transmitter 300 monitor unit 310 Receiving unit 320 decoder 330 monitor 100a Monitoring support device 150 control section 160 Storage section 161 Surveillance Support Program 170 memory 180 Communications Department 41, 42 Layer images 100-1~100-n Camera Unit 300-1~300-m monitor unit

Claims

1. a detection unit that detects specific behavior of a person in a plurality of layer images that are captured by a camera at a predetermined time and correspond to a plurality of layers that are at different distances from the camera, using a detection method corresponding to each layer; The detection unit Calculating an average value of pixel change amounts of a target corresponding to the layer in a predetermined number of layer images that correspond to the layer and are taken at different times within a predetermined time period that corresponds to the distance from the camera; The specific behavior is detected by the detection method for the layer image captured at the predetermined time point corresponding to the layer whose average value is equal to or greater than a predetermined value. Surveillance support system.

2. The computer a step of detecting a specific behavior of a person in a plurality of layer images captured by a camera at a predetermined time and corresponding to a plurality of layers at different distances from the camera, using a detection method corresponding to each layer; calculating an average value of pixel change amounts of a target corresponding to the layer in a predetermined number of layer images that correspond to the layer and are taken at different times within a predetermined time period that corresponds to the distance from the camera; detecting the specific behavior by the detection method for the layer image captured at the predetermined time point corresponding to the layer whose average value is equal to or greater than a predetermined value; A monitoring support method for:

Images