Regional safety management system based on video monitoring

Abstract

The embodiment of the invention discloses a regional safety management system based on video monitoring. A specific embodiment of the method comprises the following steps: controlling a shooting device to obtain a regional monitoring video group according to a preset shooting range; according to the brightness change in the area monitoring video group, controlling a light assembly to carry out supplementary illumination; performing dynamic feature extraction on each area monitoring video in the area monitoring video group through safety supervision equipment to generate a dynamic feature information group; adding the dynamic feature information group to a pre-generated regional static feature map to obtain a regional overall feature map group; and in response to determining that a moving target meeting a preset monitoring and early warning condition exists in the region overall feature map group, carrying out moving tracking on the moving target, and controlling a light assembly to carry out tracking illumination. According to the embodiment, occupation of storage resources can be reduced.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of area security management, and more particularly to an area security management system based on video surveillance. Background Art

[0002] Area security management refers to a technology that comprehensively manages elements such as personnel, equipment, and environment through technical means and management measures within a specific area to ensure safety and stability within the area. Currently, when performing area security management (for example, managing and maintaining the security of scenarios such as residential communities or parking lots), the commonly adopted method is to directly identify surveillance videos through a deep learning model to replace manual security monitoring and management.

[0003] However, the inventors have found that when the above method is used for area security management, the following technical problems often exist:

[0004] In order to archive the data during the security management process, it is necessary to store both the surveillance videos and the recognition results, thus requiring more storage resources.

[0005] The above information disclosed in this background art section is only used to enhance the understanding of the background of the inventive concept, and thus, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] This summary of the disclosure is used to introduce concepts in a brief form, which will be described in detail in the following detailed implementation section. This summary of the disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0007] Some embodiments of the present disclosure propose an area security management system based on video surveillance to solve the technical problems mentioned in the above background art section.

[0008] In a first aspect, some embodiments of the present disclosure provide a regional security management system based on video surveillance. The method includes: controlling the shooting device to obtain a group of regional surveillance videos according to a preset shooting range, where the shooting device includes at least one network camera; controlling the lighting component to perform supplementary lighting according to the brightness change in the group of regional surveillance videos; extracting dynamic features from each regional surveillance video in the group of regional surveillance videos through the security supervision device to generate a group of dynamic feature information, where the dynamic feature information represents the line feature or point feature of a moving object; adding the group of dynamic feature information to a pre-generated regional static feature map to obtain a group of regional overall feature maps; in response to determining that there is a moving target in the group of regional overall feature maps that meets the preset monitoring and warning conditions, performing moving tracking on the moving target, and controlling the lighting component to perform tracking lighting, where the regional overall feature map corresponding to the moving target is stored.

[0009] In a second aspect, some embodiments of the present disclosure provide a device for a regional security management system based on video surveillance. The device includes: a shooting unit configured to control a shooting device to obtain a group of regional surveillance videos according to a preset shooting range, where the shooting device includes at least one network camera; a lighting control unit configured to control a lighting component to perform supplementary lighting according to the brightness change in the group of regional surveillance videos; a dynamic feature extraction unit configured to extract dynamic features from each regional surveillance video in the group of regional surveillance videos through a security supervision device to generate a group of dynamic feature information, where the dynamic feature information represents the line feature or point feature of a moving object; a feature fusion unit configured to add the group of dynamic feature information to a pre-generated regional static feature map to obtain a group of regional overall feature maps; a target tracking unit configured to, in response to determining that there is a moving target in the group of regional overall feature maps that meets the preset monitoring and warning conditions, perform moving tracking on the moving target, and control the lighting component to perform tracking lighting, where the regional overall feature map corresponding to the moving target is stored.

[0010] In a third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more programs thereon, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method described in any implementation manner of the first aspect.

[0011] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium storing a computer program thereon, where when the program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.

[0012] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: Through the area security management system based on video surveillance in some embodiments of the present disclosure, the occupation of storage resources can be reduced. Specifically, the reason for the relatively large occupation of storage resources is that in order to archive the data in the security management process, it is necessary to store the surveillance video and the recognition results simultaneously. Based on this, in the area security management system based on video surveillance in some embodiments of the present disclosure, first, according to the preset shooting range, control the above-mentioned shooting device to obtain a group of area surveillance videos. Among them, the above-mentioned shooting device includes at least one network camera. Then, considering the influence of brightness change on the shooting effect of the camera, therefore, according to the brightness change in the above-mentioned group of area surveillance videos, control the above-mentioned lighting components to perform supplementary lighting. Thus, it is used to improve the clarity of the images captured by the camera. Furthermore, it is convenient to improve the accuracy of the recognition results. After that, through the above-mentioned security supervision device, dynamic feature extraction is performed on each area surveillance video in the above-mentioned group of area surveillance videos to generate a group of dynamic feature information, where the dynamic feature information represents the line features or point features of moving objects. Through dynamic feature extraction, the features of moving targets can be extracted from the area surveillance videos. At the same time, in combination with the area static feature map, the above-mentioned group of dynamic feature information is added to the pre-generated area static feature map to obtain a group of area overall feature maps. Thus, the accuracy of the area overall feature maps can be improved. In addition, because the pre-set area static feature map is introduced, only dynamic features need to be extracted when performing area surveillance video recognition, thereby reducing the consumption of computing resources in the video recognition process. Finally, in response to determining that there is a moving target in the above-mentioned group of area overall feature maps that meets the preset monitoring and warning conditions, perform moving tracking on the moving target, and control the above-mentioned lighting components to perform tracking lighting. Among them, the area overall feature map corresponding to the moving target is stored. Here, through moving tracking, dangerous targets can be tracked and illuminated. In addition, by storing the area overall feature maps, compared with storing all the recognition results, the occupation of storage resources can be greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In combination with the accompanying drawings and referring to the following specific embodiments, the above and other features, advantages and aspects of the various embodiments of the present disclosure will become more obvious. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic, and the elements and elements are not necessarily drawn to scale.

[0014] Figure 1 is a schematic diagram of an application scenario of the area security management system based on video surveillance according to the present disclosure;

[0015] Figure 2 is a flowchart of some embodiments of the area security management system based on video surveillance according to the present disclosure;

[0016] Figure 3 It is a schematic diagram of the front top view of the monitoring area;

[0017] Figure 4 It is a schematic structural diagram of some embodiments of the area security management system device based on video monitoring according to the present disclosure;

[0018] Figure 5 It is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed implementation manners

[0019] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0020] In addition, it should be noted that only parts related to the relevant invention are shown in the drawings for the convenience of description. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0021] It should be noted that concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules or units.

[0022] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly specified in the context, it should be understood as "one or more".

[0023] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0024] The personal information of users involved in the present disclosure (such as operations such as collection, storage, and use of user monitoring videos). Before performing the corresponding operations, relevant organizations or individuals shall fulfill obligations including conducting personal information security impact assessments, fulfilling the obligation of informing the personal information subject, and obtaining the prior authorization and consent of the personal information subject.

[0025] The present disclosure will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0026] Figure 1It is a schematic diagram of an application scenario of a regional security management system based on video surveillance according to some embodiments of the present disclosure.

[0027] In Figure 1 the application scenario, the above-mentioned regional security management system 101 may include: a shooting device 102, a security supervision device 103, and a lighting component 104. First, the regional security management system 101 may control the above-mentioned shooting device 102 to obtain a group of regional surveillance videos according to a preset shooting range. Among them, the above-mentioned shooting device includes at least one network camera. Then, the regional security management system 101 controls the above-mentioned lighting component 104 to perform supplementary lighting according to the brightness change in the above-mentioned group of regional surveillance videos. After that, the regional security management system 101 may extract dynamic features from each regional surveillance video in the above-mentioned group of regional surveillance videos through the above-mentioned security supervision device 103 to generate a group of dynamic feature information. Among them, the dynamic feature information represents the line feature or point feature of a moving object. Then, the above-mentioned group of dynamic feature information is added to a pre-generated regional static feature map to obtain a group of regional overall feature maps. Finally, the regional security management system 101 may, in response to determining that there is a moving target in the above-mentioned group of regional overall feature maps that meets the preset monitoring and warning conditions, perform moving tracking on the moving target and control the above-mentioned lighting component 104 to perform tracking lighting.

[0028] It should be noted that the above-mentioned regional security management system 101 may be hardware or software. When the regional security management system is hardware, it may be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the regional security management system is embodied as software, it may be installed in the above-mentioned listed hardware devices. It may be implemented as, for example, multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here. It should be understood that Figure 1 the number of regional security management systems in

[0029] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0030] Figure 2 Flow 200 according to some embodiments of the regional security management system based on video surveillance of the present disclosure is shown. The regional security management system based on video surveillance includes the following steps:

[0031] Step 201, control a shooting device to obtain a group of regional surveillance videos according to a preset shooting range.

[0032] In some embodiments, the area security management system based on video surveillance can control the above-mentioned shooting device to obtain a group of area surveillance videos by wired or wireless means according to a preset shooting range. Among them, the above-mentioned shooting device includes at least one network camera. The shooting range can be the area that the camera can shoot. Here, each network camera corresponds to an area surveillance video.

[0033] In some optional implementation manners of some embodiments, at least one network camera in the above-mentioned shooting device is arranged at different positions in the shooting area. And the above-mentioned execution subject controls the above-mentioned shooting device to obtain a group of area surveillance videos according to a preset shooting range, including:

[0034] First step, within the above-mentioned shooting range, allocate shooting areas for the above-mentioned at least one network camera, where the above-mentioned shooting range fully covers the above-mentioned shooting areas. Here, the camera position coordinates and shooting angles pre-specified for each network camera can be obtained. Then, the shooting areas within the shooting range can be allocated to the network cameras through the camera position coordinates and shooting angles.

[0035] Second step, control the above-mentioned at least one network camera to rotate and shoot to shoot the allocated shooting areas, and obtain a group of area surveillance videos. Among them, the network camera is a rotatable camera, so that each network camera can be controlled to rotate and shoot to shoot the allocated shooting areas, and obtain a group of area surveillance videos. For example, the rotatable cameras can include: spherical cameras, gun-type cameras, etc.

[0036] Optionally, if the network camera is a fixed-area camera, it does not perform rotational shooting.

[0037] Step 202, control the lighting component to perform supplementary lighting according to the brightness change in the area surveillance video group.

[0038] In some embodiments, the above-mentioned execution subject can control the above-mentioned lighting component to perform supplementary lighting according to the brightness change in the above-mentioned area surveillance video group.

[0039] In practice, considering that the shooting environment is prone to brightness changes, such as rainy weather, night scenes, etc. Specifically, for example, the brightness of the scene under the tree is lower than that of the unobstructed scene. Although infrared light can be used to assist shooting, by providing supplementary lighting, the video clarity can be further improved.

[0040] In some optional implementation manners of some embodiments, the above-mentioned lighting component includes a group of street lamp devices, a wall strip light component, and a group of spotlight devices; and the above-mentioned execution subject controls the above-mentioned lighting component to perform supplementary lighting according to the brightness change in the above-mentioned area surveillance video group, including:

[0041] First step, perform video brightness detection on each area monitoring video in the above area monitoring video group to obtain a group of real-time video brightness detection values. Among them, the video brightness of each area monitoring video in the above area monitoring video group can be detected through a brightness detection algorithm to obtain real-time video brightness detection values. Here, each area monitoring video can correspond to a real-time video brightness detection value.

[0042] As an example, the brightness detection algorithm can include but is not limited to at least one of the following: mean method, histogram method, logarithmic transformation algorithm, local binary pattern algorithm, etc.

[0043] Second step, in response to determining that there is a real-time video brightness detection value less than the brightness threshold in the above group of real-time video brightness detection values, mark the corresponding network camera. Among them, the network camera with a real-time video brightness detection value less than the brightness threshold can be marked. For example, adjust the brightness identifier corresponding to the network camera from 0 to 1 to indicate low brightness.

[0044] Third step, in the pre-generated top view of the monitoring area, perform brightness area marking on the area monitoring video corresponding to the marked network camera to generate a marked area top view. Among them, the positions and lighting areas of various lighting devices are marked in the above top view of the monitoring area. Here, the top view of the monitoring area can be established from the top view of the area photographed by a drone in advance, or can be established by scanning area point cloud data, photographing area images, etc. During the establishment process, for fixed objects in the photographed image (such as buildings, power distribution equipment, trees, roads, flower beds, the position coordinates of network cameras, the position coordinates of lighting devices, etc.), after identification, they can be represented in a simplified symbol way. For example, fixed objects are represented by coordinate points or lines. The position coordinates of each fixed object in the top view of the monitoring area correspond one by one to the objects in the actual photographed area. In addition, the brightness area identifier can be to perform brightness area marking on the area photographed by the marked network camera in the top view of the monitoring area. Specifically, first, the position of the corresponding network camera in the top view of the monitoring area can be determined according to the identifier of the marked network camera. Then, the photographed area corresponding to the network camera is marked to obtain a marked area top view.

[0045] As an example, as Figure 3 shown in the schematic diagram of the top view of the monitoring area. In the figure, the position coordinates of the network camera, the corresponding photographed area, and the position distribution of fixed objects (such as buildings, flower beds, street lights) are marked in a minimalist way (that is, using lines to represent fixed areas and using points or symbols to represent fixed objects). Figure 3The triangles at the four corners represent the rotating cameras, and the cylinders represent the fixed cameras. The black dots represent the position coordinates of the lighting devices. The circular area is the pond. The rectangular area is the building.

[0046] Here, on the display terminal, if it is necessary to display a more detailed top view of the monitoring area (for example, a top view of the monitoring area with simulation images), different symbol markings in the top view of the monitoring area can be associated with the corresponding rendering models in advance. Thus, when it is necessary to display a more detailed top view of the monitoring area, the corresponding rendering method can be selected (for example, select the option of "display building"). As a result, the corresponding top view of the monitoring area can be displayed. In this way, the offline rendering display method reduces the occupation of real-time computing resources and the occupation of storage resources.

[0047] In practice, since the top view of the monitoring area is a grayscale image generated by simplified symbols, the color depth and the number of color channels of the image can be greatly reduced. Thus, subsequent storage can reduce the occupation of storage resources.

[0048] Optionally, when the display terminal displays the top view of the marked area, the captured area of the mark can be highlighted, or the captured area can be marked by highlighting, etc. In addition, the top view of the monitoring area can be fixed. Therefore, in order to further reduce its occupation of storage resources. The marks in the top view of the monitoring area can also be rendered and removed by options. For example, in the control options of the above-mentioned top view of the monitoring area on the display terminal, check the option of "display lighting area", and then the lighting area corresponding to each lighting device can be rendered and marked in the top view of the monitoring area. When not checked, the marking of the lighting area in the top view of the monitoring area can be eliminated. Thus, the occupation of storage resources is reduced.

[0049] Fourth step, control the street lamp devices or fence light strips corresponding to the lighting areas that coincide with the marked brightness areas in the top view of the marked area to be turned on to supplement the lighting of the captured area of the marked network camera. Among them, first, the lighting areas that coincide with the marked brightness areas can be determined, and the identifiers of the street lamp devices or fence light strips with brightness association can be determined accordingly. Then, the corresponding street lamp devices or fence light strips can be controlled to be turned on through the identifiers to supplement the lighting of the captured area of the marked network camera.

[0050] In practice, by introducing safety supervision devices, the following steps can be taken to identify the characteristics of moving targets (such as pedestrians) to improve the safety within the area.

[0051] Step 203: Use the security supervision device to extract dynamic features from each regional surveillance video in the regional surveillance video group to generate a dynamic feature information group.

[0052] In some embodiments, the above-mentioned execution entity may use the above-mentioned security supervision device to extract dynamic features from each regional surveillance video in the above-mentioned regional surveillance video group to generate a dynamic feature information group. Among them, the dynamic feature information characterizes the line feature or point feature of the moving object. Here, the line feature may be the detection frame of the detected moving object. The point feature may be the position coordinate point of the detected moving object.

[0053] In some optional implementation manners of some embodiments, the above-mentioned execution entity uses the above-mentioned security supervision device to extract dynamic features from each regional surveillance video in the above-mentioned regional surveillance video group to generate a dynamic feature information group, including:

[0054] First step: Detect moving objects in the video images of each regional surveillance video in the above-mentioned regional surveillance video group to mark the moving objects in the inactive area and obtain a moving object identification group. Among them, each moving object identification corresponds to one or more moving objects with close distances. For each newly generated moving object identification, establish a corresponding observation queue. For the moving object identification that exceeds the shooting area, recycle the corresponding observation queue. Here, the observation queue can be an array with a fixed-size structure or a queue with a list structure whose size can be dynamically adjusted. Specifically, a preset detection algorithm can be used to detect moving objects in the video images of each regional surveillance video in the above-mentioned regional surveillance video group to mark the moving objects in the inactive area and obtain a moving object identification group. Secondly, each moving object identification may correspond to a moving object inspection frame. In addition, each newly detected moving object can be regarded as a whole, that is, establish the same observation queue for observation. If the distance between multiple moving objects in the same observation queue is greater than the preset distance threshold, the observation queue can be split and corresponding observation queues can be established respectively. This ensures the accuracy of the description of the moving object intention in each subsequent observation queue.

[0055] As an example, the detection algorithm may include, but is not limited to, at least one of the following: R-CNN (Regions with CNN features): Region Convolutional Neural Network Features, Fast R-CNN: Fast Region Convolutional Neural Network, YOLO (You Only Look Once) object detection algorithm, SSD (Single Shot MultiBox Detector) Single Shot MultiBox Detector.

[0056] Step 2: For each moving target identifier in the above moving target identifier group and the corresponding observation queue, perform the following dynamic feature extraction steps:

[0057] Step 1: Monitor the dangerous behaviors of the moving targets corresponding to the above moving target identifiers to obtain dangerous behavior monitoring information. The dangerous behavior monitoring information includes: moving target position coordinates, dangerous behavior identifiers, and corresponding static associated object position identifiers. First, through coordinate conversion, the midpoint coordinate of the lower edge of the moving target detection frame corresponding to the above moving target identifier can be converted from the image coordinate system of the video image to the orthographic top view of the monitoring area to obtain the converted moving target coordinates. Then, with the above converted moving target coordinates as the center and using a preset radius, mark the moving target safety area. After that, the distance values between each fixed object in the moving target safety area and the center (i.e., the converted moving target coordinates) can be determined. If the distance value is less than or equal to the preset threshold and the duration is greater than the preset duration threshold, the corresponding fixed object coordinates can be determined as the static associated object position identifiers. After that, the above converted moving target coordinates are determined as the moving target position coordinates. Finally, select the dangerous behavior identifier corresponding to the above static associated object position identifier from the pre-established dangerous behavior table. In addition, if the distance between the converted moving target coordinates and the coordinates of the fixed object is greater than the preset threshold, or the duration is less than or equal to the preset duration threshold, an empty dangerous behavior identifier is generated.

[0058] As an example, different fixed objects (i.e., static associated objects) can correspond to different dangerous behavior identifiers. For example, if it is detected that a moving target is close to power distribution equipment for a long time (such as a child playing near or damaging power distribution equipment), the identifier representing "electric shock danger" can be determined as the corresponding dangerous behavior identifier a. Another example is that if it is detected that a moving target is close to the edge of a pond for a long time, the identifier representing "falling into water danger" can be determined as the corresponding dangerous behavior identifier b. Another example is that if it is detected that a moving target is climbing over a wall, the identifier representing "falling danger" can be determined as the corresponding dangerous behavior identifier c.

[0059] Step 2: In response to detecting that the dangerous behavior monitoring information meets the preset dangerous behavior conditions, use the above-mentioned observation queue to mark the moving target interval video in the corresponding area monitoring video, and use the above-mentioned dangerous action monitoring information to determine the moving path corresponding to the moving target identifier. Among them, the above-mentioned dangerous behavior condition is that the dangerous behavior identifier included in the dangerous behavior monitoring information is not empty. Secondly, each timestamp in the observation queue where the moving target is located can be determined as the observation time period. Mark the moving target interval video in the corresponding area monitoring video for the above-mentioned observation time period. Here, marking the moving target interval video can be used to store and record the dangerous behavior of the moving target in a lossless compression manner for subsequent extraction and playback. For videos in other time periods, a lossy compression method can be used to increase the compression ratio and reduce the occupation of storage resources. Then, the position coordinates of the moving target in consecutive time periods can be extracted from the above-mentioned observation queue to obtain a sequence of position coordinates. Finally, the above-mentioned sequence of position coordinates can be used as the moving path.

[0060] Step 3: According to the above-mentioned moving path, perform moving target intention recognition on the above-mentioned moving target interval video to generate moving target intention description information, where the above-mentioned moving target intention description information represents the moving intention of the moving target in the monitoring area. For example, intentions such as climbing over (the wall), squatting, lying down, climbing (the tree), etc.

[0061] Here, the initial coordinates and the termination coordinates of the above-mentioned moving path can be extracted. Then, through a preset path planning algorithm, determine the shortest path between the above-mentioned initial coordinates and the termination coordinates. In addition, when planning the path, the coordinates of fixed objects can be introduced as obstacles to avoid obstacles. After that, through a preset path similarity algorithm, determine the path similarity between the above-mentioned shortest path and the above-mentioned moving path. Here, considering that the moving target may have a clear or unclear target, which leads to a relatively simple moving path when the target is clear. When the target is unclear, there is a situation of small-range repeated movement, resulting in a relatively high degree of repetition and complexity of the generated moving path. Therefore, by generating the path similarity, it can be used to determine whether the moving intention of the moving target is clear.

[0062] As an example, the above-mentioned path planning algorithm may include, but is not limited to, at least one of the following: A* algorithm, Dijkstra's algorithm, breadth-first search algorithm, etc.

[0063] Then, considering that the paths of moving targets are highly repetitive and tortuous, common path similarity algorithms (e.g., Euclidean distance, dynamic time warping, etc.) are prone to ignoring the corresponding relationship between two paths on the time axis, resulting in misjudging the similarity of the two paths. Therefore, a method based on topological structure, i.e., path graph analysis, or a method based on path shape (i.e., polygon approximation algorithm: Douglas-Peucker algorithm) can be used as the path similarity algorithm. Thus, the time order of different coordinates on the path can be combined to compare the path similarity. When generating the similarity, the case where the path has a U-turn can be combined to compare the path similarity. Thereby, the accuracy of the path similarity can be improved.

[0064] After that, for a moving path with unclear moving intention, a lightweight human pose recognition algorithm can also be used to perform sampling-based moving target pose recognition on the video images corresponding to each position coordinate on the moving path, obtaining a moving target pose identification sequence. Here, the sampling can be performed at a preset time interval, or the moving target pose recognition can be performed on the video images corresponding to the inflection point positions on the moving path. For example, the moving target pose identification can be used to represent at least one of the following postures: walking, wandering, climbing over, running, jumping, crawling, squatting, turning around, waving, using a mobile phone, lying down, etc. At the same time, for a moving path with clear moving intention, the video image corresponding to the last position coordinate on the moving path can be recognized through the above lightweight human pose recognition algorithm, obtaining moving target intention description information. Thus, through the lightweight algorithm, the computing power consumption can be reduced when identifying the intention description of the moving target.

[0065] As an example, the moving target intention description information can be: the moving target is climbing over a wall, or the moving target is climbing a tree, etc.

[0066] After that, the continuous moving target pose identification sequence is input into a pre-trained decision tree model to output the moving intention identification of the moving target. Here, the decision tree model can be trained through a publicly available dataset obtained in advance (e.g., NTU RGB+D dataset, Kinetics human action video dataset). In addition, during the training process, it is only necessary to train the decision tree to classify the identification of the moving target's behavior intention based on the continuous moving target pose identification. Here, the behavior intention of the moving target can include at least one of the following: climbing over (a wall), squatting, lying down, crawling (a tree), playing in the water, etc. For the continuous moving target pose identification that is difficult to distinguish, it can be continuously observed, or the moving intention identification of "wandering" can be output. Finally, according to the preset description sentence pattern, the moving intention identification can be substituted to obtain the moving target intention description information.

[0067] Here, the lightweight human pose recognition algorithm may include, but is not limited to, at least one of the following: LightweightOpenPose, lightweight human pose estimation model, CenterNet (Object Keypoint Similarity) center keypoint pose estimation network, or PoseCNN pose convolutional neural network, etc.

[0068] In practice, for the generation of the intention description information of a moving target, if the method of completely using a deep learning model for recognition is adopted, a relatively complex model structure needs to be designed. At the same time, it also requires high computing power support. To further reduce the computing power, the above steps first detect whether there are dangerous behavior identifiers and corresponding associated static associated object position identifiers of the moving target through the method of dangerous behavior detection. Specifically, by determining the moving path of the moving target, the intention of the moving target is judged. After that, only by combining the lightweight human pose recognition algorithm with the decision tree model can the generation of the intention description information of the moving target be replaced by completely using the deep learning model. Here, because the static associated object position identifier is introduced, it can be used as a fixed spatial feature to define the dangerous behavior of the moving target. Thus, not only can the deficiency of the deep learning model in scene depth recognition be made up for, but also the dangerous behavior of the moving target can be further assisted in discrimination. Thereby, the computing power consumption is reduced.

[0069] Step four, determine the above-mentioned dangerous position monitoring information, the above-mentioned moving path, and the above-mentioned moving target intention description information as the dynamic feature information in the above-mentioned dynamic feature information group.

[0070] Step 204, add the dynamic feature information group to the pre-generated regional static feature map to obtain a group of regional overall feature maps.

[0071] In some embodiments, the above-mentioned execution subject may add the above-mentioned dynamic feature information group to the pre-generated regional static feature map to obtain a group of regional overall feature maps.

[0072] In some optional implementation manners of some embodiments, the above-mentioned execution subject adds the above-mentioned dynamic feature information group to the pre-generated regional static feature map to obtain a group of regional overall feature maps, including:

[0073] Add the movement paths, movement target position coordinates, dangerous behavior identifiers, and corresponding static associated object position identifiers included in each dynamic feature information in the above dynamic feature information group to the above regional static feature map to obtain a group of regional overall feature maps. Among them, the above regional static feature map can be a pre-generated top view of the monitored area or a captured area map established from the shooting perspective of the corresponding network camera. The regional static feature map may include the position coordinates of fixed objects. Here, the captured area map established from the shooting perspective of the corresponding network camera is an image established from the camera perspective. Thus, the regional features can be reflected from the side angle. The top view of the monitored area can reflect the regional features from the top view angle. Secondly, the movement path can be added to the top view of the monitored area according to the coordinates of the movement path. For the dangerous behavior identifier and the corresponding static associated object position identifier, the corresponding icons can be added to the corresponding positions.

[0074] Step 205, in response to determining that there is a moving target in the above group of regional overall feature maps that meets the preset monitoring and warning conditions, perform moving tracking on the moving target and control the above lighting component for tracking illumination.

[0075] In some embodiments, the above execution entity may, in response to determining that there is a moving target in the above group of regional overall feature maps that meets the preset monitoring and warning conditions, perform moving tracking on the moving target and control the above lighting component for tracking illumination. Among them, the monitoring and warning conditions may be keywords including dangerous behavior descriptions in the moving target intention description information. For example, keywords such as climbing over the wall, getting into the water, climbing a tree, fiddling with the distribution box, etc.

[0076] In some optional implementation manners of some embodiments, the above execution entity, in response to determining that there is a moving target in the above group of regional overall feature maps that meets the preset monitoring and warning conditions, performs moving tracking on the moving target, including:

[0077] In the first step, in response to determining that there are dangerous behavior identifiers and corresponding static associated object position identifiers in the regional overall feature map in the above group of regional overall feature maps, it is determined that the regional overall feature map meets the preset monitoring and warning conditions.

[0078] In practice, the dangerous behavior identifiers may include: climbing over, squatting, climbing up, rummaging through, etc. The corresponding static associated object position identifiers may include: walls, ponds, trees, power distribution equipment, etc. Therefore, the corresponding moving target intention description information is described as keywords such as climbing over the wall, getting into the water, climbing a tree, fiddling with the distribution box, etc. Thus, the above monitoring and warning conditions are triggered.

[0079] Second step, in response to determining that the moving target position coordinates of the moving target that meets the above monitoring and early warning conditions exceed the orientation range of the corresponding network camera, control the above network camera to perform moving tracking on the above moving target. Among them, the moving target position coordinates exceeding the orientation range of the corresponding network camera indicates that the position of the moving target exceeds the current shooting range of the network camera. In practice, for a steerable network camera, if the tracking target exceeds the current shooting range, the shooting direction can be rotated to track the moving target.

[0080] Third step, in response to determining that the above moving target exceeds the shooting range of the above network camera, call the adjacent network camera to perform linkage tracking on the above moving target according to the moving path corresponding to the above moving target. Among them, the moving target exceeding the shooting range of the above network camera indicates that the position of the above moving target exceeds all shooting areas of the above network camera. In practice, the moving target exceeding the shooting range of the above network camera means that the moving target is blocked or exceeds all shooting ranges of the above network camera. Therefore, the adjacent network camera can be called to perform linkage tracking on the above moving target. Here, for different camera shooting areas, their adjacent relationships can be established in advance. For example, the camera labeled s1 and the camera s2 are adjacent to each other on the left and right. If the target moves to the right and exceeds the shooting area within the s1 lens, the s2 camera can be called as the adjacent network camera to rotate the shooting direction to perform linkage tracking on the above moving target. In addition, during the linkage tracking process, the shooting data corresponding to the same moving target (for example, video images, position coordinates of the moving target, etc.) can correspond to the same observation queue.

[0081] In some optional implementation manners of some embodiments, the above execution subject controls the above lighting component to perform tracking lighting, including:

[0082] In response to determining that the current time point is a night time period or the real-time video brightness detection value is less than the brightness threshold, turn on the spotlight device that is closest to the above moving target and not blocked, and control the spotlight device to perform tracking lighting on the moving target position coordinates where the above moving target is located. Among them, the distance between the spotlight device and the above moving target is determined to be the closest distance and whether it is blocked through the position coordinates and lighting area of the spotlight device marked in the orthographic top view of the above monitoring area. Here, by judging whether the distance between the spotlight device and the above moving target is the closest distance and whether it is blocked, the corresponding spotlight device can be selected. Then the spotlight device can be controlled to perform tracking lighting on the moving target position coordinates where the above moving target is located.

[0083] In practice, considering that dangerous behaviors occur in the monitoring area during the night time, if voice prompts are used, it is easy to interfere with other users. If security guards are prompted to conduct inspections, it is easy to miss the time. Therefore, by introducing spotlight devices for tracking illumination, it is possible to more directly cause moving targets with the intention of performing dangerous actions to change their intentions. Thus, not only can a bright and safe scene be provided for the moving targets, but also a warning effect can be achieved. Here, if the moving target is identified as an animal, only security guards can be prompted to conduct inspections. Additionally, if a moving target showing dangerous behavior is detected during the day, reminders can be given through the installed audio playback devices.

[0084] Optionally, after performing the reminder operation, the video image can be tracked and recognized again through the human body posture recognition algorithm. If it is still performing dangerous actions, such as completing dangerous behaviors like climbing over a wall, playing in the water, climbing a tree, or fiddling with electrical distribution equipment. Then the video of the moving target interval corresponding to its observation queue can be sent to the regional display screen for video carousel. Thus, the purpose of warning can be achieved.

[0085] The above embodiments of the present disclosure have the following beneficial effects: Through the area security management system based on video surveillance according to some embodiments of the present disclosure, the occupation of storage resources can be reduced. Specifically, the reason for the relatively large occupation of storage resources is that in order to archive the data during the security management process, it is necessary to store both the surveillance videos and the recognition results at the same time. Based on this, in the area security management system based on video surveillance according to some embodiments of the present disclosure, first, according to the preset shooting range, the above-mentioned shooting device is controlled to obtain a group of area surveillance videos. Among them, the above-mentioned shooting device includes at least one network camera. Then, considering the influence of brightness changes on the shooting effect of the camera, the above-mentioned lighting component is controlled to perform supplementary lighting according to the brightness changes in the above-mentioned group of area surveillance videos. Thereby, it is used to improve the clarity of the images captured by the camera. Furthermore, it is convenient to improve the accuracy of the recognition results. After that, through the above-mentioned security supervision device, dynamic feature extraction is performed on each area surveillance video in the above-mentioned group of area surveillance videos to generate a group of dynamic feature information, where the dynamic feature information characterizes the line features or point features of moving objects. Through dynamic feature extraction, the features of moving targets can be extracted from the area surveillance videos. At the same time, in combination with the area static feature map, the above-mentioned group of dynamic feature information is added to the pre-generated area static feature map to obtain a group of area overall feature maps. Thus, the accuracy of the area overall feature maps can be improved. In addition, because the pre-set area static feature map is introduced, only dynamic features need to be extracted during area surveillance video recognition, thereby reducing the consumption of computing resources during the video recognition process. Finally, in response to determining that there is a moving target in the above-mentioned group of area overall feature maps that meets the preset monitoring and warning conditions, mobile tracking is performed on the moving target, and the above-mentioned lighting component is controlled to perform tracking lighting. Among them, the area overall feature map corresponding to the moving target is stored. Here, through mobile tracking, dangerous targets can be tracked and illuminated. In addition, by storing the area overall feature maps, compared with storing all the recognition results, the occupation of storage resources can be greatly reduced.

[0086] Further referring to Figure 4 , as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of an area security management device based on video surveillance. These device embodiments correspond to Figure 2 the method embodiments shown, and the area security management device based on video surveillance can be specifically applied to various electronic devices.

[0087] As shown in Figure 4As shown, the area security management device 200 based on video surveillance in some embodiments includes: a shooting unit 401, a lighting control unit 402, a dynamic feature extraction unit 403, a feature fusion unit 404, and a target tracking unit 405. Among them, the shooting unit 401 is configured to control a shooting device to obtain a group of area surveillance videos according to a preset shooting range, where the shooting device includes at least one network camera; the lighting control unit 402 is configured to control a lighting component to perform supplementary lighting according to the brightness change in the group of area surveillance videos; the dynamic feature extraction unit 403 is configured to perform dynamic feature extraction on each area surveillance video in the group of area surveillance videos through a security supervision device to generate a group of dynamic feature information, where the dynamic feature information characterizes the line feature or point feature of a moving object; the feature fusion unit 404 is configured to add the group of dynamic feature information to a pre-generated area static feature map to obtain a group of area overall feature maps; the target tracking unit 405 is configured to, in response to determining that there is a moving target in the group of area overall feature maps that meets a preset monitoring and warning condition, perform moving tracking on the moving target, and control the lighting component to perform tracking lighting, where the area overall feature map corresponding to the moving target is stored.

[0088] It can be understood that the various units described in the area security management device 400 based on video surveillance correspond to the respective steps in the method described in the reference Figure 2 description. Thus, the operations, features, and beneficial effects described above for the method also apply to the area security management device 400 based on video surveillance and the units included therein, and will not be elaborated herein.

[0089] Next, refer to Figure 5 , which shows a schematic structural diagram of an electronic device (area security management system) suitable for implementing some embodiments of the present disclosure. Figure 5 The electronic device shown is merely an example and should not impose any limitation on the functions and usage scope of the embodiments of the present disclosure. As Figure 5 shown, the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory. The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the front-end page monitoring methods. The processor is used to provide computing and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium, and when the computer program is executed by the processor, the processor can execute any one of the front-end page monitoring methods. The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand, Figure 5The structure shown is only a block diagram of some of the structures related to the present disclosure solution, and does not constitute a limitation on the computer device to which the present disclosure solution is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0090] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0091] Among them, in one embodiment, the above-mentioned processor is used to run the computer program stored in the memory to implement the following steps: controlling the above-mentioned photographing device to obtain a regional monitoring video group according to a preset photographing range, where the above-mentioned photographing device includes at least one network camera; controlling the above-mentioned lighting component to perform supplementary lighting according to the brightness change in the above-mentioned regional monitoring video group; extracting dynamic features from each regional monitoring video in the above-mentioned regional monitoring video group through the above-mentioned safety supervision device to generate a dynamic feature information group, where the dynamic feature information characterizes the line feature or point feature of a moving object; adding the above-mentioned dynamic feature information group to a pre-generated regional static feature map to obtain a regional overall feature map group; in response to determining that there is a moving target in the above-mentioned regional overall feature map group that meets the preset monitoring and early warning conditions, performing moving tracking on the moving target, and controlling the above-mentioned lighting component to perform tracking lighting, where the regional overall feature map corresponding to the moving target is stored.

[0092] The present disclosure embodiment also provides a computer-readable storage medium, on which a computer program is stored, and the computer program includes program instructions. The method implemented when the program instructions are executed may refer to the various embodiments of the regional security management system based on video monitoring of the present disclosure.

[0093] Among them, the above computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The above computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a SmartMedia Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device.

[0094] It should be noted that in this document, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or system including a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or system. Without further limitation, an element defined by the phrase "including one..." does not exclude the presence of additional identical elements in the process, method, article or system including such element.

[0095] The above description is only some preferred embodiments of the present disclosure and an explanation of the technical principles applied. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) disclosed in the embodiments of the present disclosure that have similar functions.

Claims

1. A regional security management system based on video surveillance, characterized in that: The regional safety management system includes: a shooting device, a safety monitoring device and a lighting component, wherein: According to a preset shooting range, controlling the shooting device to obtain a regional monitoring video group, wherein the shooting device includes at least one network camera; Controlling the light assembly to provide supplementary lighting according to brightness changes in the regional monitoring video group; By means of the security monitoring device, dynamic feature extraction is performed on each regional monitoring video in the regional monitoring video group to generate a dynamic feature information group, wherein the dynamic feature information represents line features or point features of a moving object; Adding the dynamic feature information group to a pre-generated regional static feature map to obtain a regional overall feature map group; In response to determining that there is a mobile target that meets the preset monitoring and warning conditions in the regional overall feature map group, the mobile target is tracked, and the lighting assembly is controlled to perform tracking lighting, wherein the regional overall feature map corresponding to the mobile target is stored.

2. The method according to claim 1, characterized in that The at least one network camera included in the shooting device is arranged at different positions in the shooting area; as well as The step of controlling the shooting device to obtain a regional monitoring video group according to a preset shooting range includes: Within the shooting range, allocating a shooting area to the at least one network camera, wherein the shooting range fully covers the shooting area; The at least one network camera is controlled to rotate and shoot to shoot the assigned shooting area, and a regional monitoring video group is obtained.

3. The method according to claim 1, characterized in that: The lighting assembly includes a street light device group, a wall light strip assembly and a spotlight device group; as well as The step of controlling the light assembly to perform supplementary lighting according to the brightness change in the regional monitoring video group includes: Performing video brightness detection on each regional monitoring video in the regional monitoring video group to obtain a video real-time brightness detection value group; In response to determining that there is a video real-time brightness detection value less than the brightness threshold in the video real-time brightness detection value group, marking the corresponding network camera; In the pre-generated top view of the monitoring area, the brightness area of ​​the area monitoring video corresponding to the marked network camera is marked to generate a marked area top view, wherein the position coordinates and lighting areas of various lighting equipment are marked in the top view of the monitoring area; The street lamp equipment or wall light strip corresponding to the lighting area overlapping with the marked brightness area in the top view of the marked area is controlled to be turned on, so as to provide supplementary lighting for the shooting area of ​​the marked network camera.

4. The method according to claim 3, characterized in that The method of extracting dynamic features from each regional surveillance video in the regional surveillance video group by the security supervision device to generate a dynamic feature information group includes: Performing mobile target detection on the video images in each regional surveillance video in the regional surveillance video group to mark the mobile targets in the inactive area, and obtaining a mobile target identification group, wherein each mobile target identification corresponds to one or more mobile targets at a similar distance, and for each newly generated mobile target identification, establishing a corresponding observation queue, and for the mobile target identification beyond the shooting area, reclaiming the corresponding observation queue; For each mobile target identifier in the mobile target identifier group and the corresponding observation queue, the following dynamic feature extraction steps are performed: Performing dangerous behavior monitoring on the mobile target corresponding to the mobile target identifier to obtain dangerous behavior monitoring information, wherein the dangerous behavior monitoring information includes: the mobile target position coordinates, the dangerous behavior identifier and the corresponding static associated object position identifier; In response to monitoring that the dangerous behavior monitoring information meets the preset dangerous behavior condition, using the observation queue to mark the moving target interval video in the corresponding regional monitoring video, and using the dangerous action monitoring information to determine the moving path corresponding to the moving target identifier; According to the moving path, performing moving target intention recognition on the moving target interval video to generate moving target intention description information; The dangerous location monitoring information, the moving path and the moving target intention description information are determined as dynamic feature information in the dynamic feature information group.

5. The method according to claim 4, characterized in that The step of adding the dynamic feature information group to a pre-generated regional static feature map to obtain a regional overall feature map group includes: The moving path, moving target position coordinates, dangerous behavior identification and corresponding static associated object position identification included in each dynamic feature information in the dynamic feature information group are added to the regional static feature map to obtain the regional overall feature map group, wherein the regional static feature map is a pre-generated overhead view of the monitoring area, or a shooting area map established with the shooting angle of the corresponding network camera, and the regional static feature map includes the position coordinates of fixed objects.

6. The method according to claim 5, characterized in that In response to determining that the group of regional overall characteristic graphs contains a mobile target that meets a preset monitoring and early warning condition, tracking the mobile target includes: In response to determining that a dangerous behavior identifier and a corresponding static associated object position identifier exist in a regional overall feature map in the regional overall feature map group, determining that the regional overall feature map meets a preset monitoring and warning condition; In response to determining that the moving target position coordinates of the moving target that meets the monitoring warning condition are beyond the direction range of the corresponding network camera, controlling the network camera to track the movement of the moving target, wherein the moving target position coordinates exceeding the direction range of the corresponding network camera represent that the position of the moving target is beyond the current shooting range of the network camera; In response to determining that the moving target is beyond the shooting range of the network camera, an adjacent network camera is called to perform linkage tracking on the moving target according to the moving path corresponding to the moving target, wherein the moving target being beyond the shooting range of the network camera indicates that the position of the moving target is beyond the entire shooting area of ​​the network camera.

7. The method according to claim 6, characterized in that The controlling the light assembly to perform tracking lighting comprises: In response to determining that the current time point is a night time period or the real-time brightness detection value of the video is less than a brightness threshold, turn on a spotlight device that is closest to the moving target and is not obstructed, and control the spotlight device to track and illuminate the moving target position coordinates where the moving target is located, wherein the position coordinates and lighting area of ​​the spotlight device marked in the overhead view of the monitoring area are used to determine whether the distance between the spotlight device and the moving target is the shortest distance and whether it is obstructed.

8. A regional security management device based on video surveillance, comprising: A shooting unit is configured to control a shooting device to obtain a regional monitoring video group according to a preset shooting range, wherein the shooting device includes at least one network camera; A lighting control unit configured to control the lighting assembly to provide supplementary lighting according to brightness changes in the regional monitoring video group; A dynamic feature extraction unit is configured to extract dynamic features from each of the regional surveillance videos in the regional surveillance video group through a security monitoring device to generate a dynamic feature information group, wherein the dynamic feature information represents a line feature or a point feature of a moving object; A feature fusion unit is configured to add the dynamic feature information group to a pre-generated regional static feature map to obtain a regional overall feature map group; The target tracking unit is configured to track the mobile target in response to determining that the regional overall feature map group contains a mobile target that meets preset monitoring and early warning conditions, and control the lighting assembly to perform tracking lighting, wherein the regional overall feature map corresponding to the mobile target is stored.

9. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images