Security risk monitoring and early warning method and system based on digital earth map

Abstract

The invention belongs to the technical field of monitoring, and discloses a safety risk monitoring and early warning method and system based on a digital earth map, and the method comprises the steps: obtaining pedestrian video information monitored by a plurality of sampling points in a preset region, and determining the motion track of each pedestrian based on the pedestrian video information; integrating the movement tracks of the pedestrians in a pre-constructed digital earth map of a preset area to obtain a track digital map, dividing the digital earth map into a plurality of map areas, and visually displaying the track digital map in each map area; determining a congestion density within each map area in a future period based on the trajectory digital map; based on the crowding density in each map area in the future time period, constructing a people flow safety level of each map area; and when the people flow safety level of any map area reaches the risk level, generating risk early warning information. The method has a good display effect and can accurately predict the people flow of the whole region.

Description

Technical Field

[0001] The present invention belongs to the technical field of monitoring, and particularly relates to a safety risk monitoring and early warning method and system based on a digital earth map. Background Art

[0002] In real life, humans often participate in various social activities in groups, such as cultural and sports performances, transportation, shopping and leisure, and tourism visits. When humans participate in public social activities, in order to maintain the safety of the activity site, real-time crowd monitoring and early warning are particularly important, which can effectively avoid congestion and chaos at the activity site and lead to stampede accidents.

[0003] Crowd monitoring is a technical means to monitor and analyze the number and flow of pedestrians in a specific area to evaluate the degree of crowd density. Traditional crowd monitoring methods mainly rely on manual statistics and some simple technical means, such as manual counting, infrared detection, and ultrasonic detection.

[0004] Among them, manual counting: Count the pedestrians passing through a certain area manually at a specific time point. This method has low accuracy and is greatly affected by human factors.

[0005] Among them, infrared detection: Detect the human body heat radiation by using an infrared sensor, and estimate the number of pedestrians by calculating the change in the heat radiation received by the sensor. This method is greatly affected by obstacles.

[0006] Among them, ultrasonic detection: Calculate the distance between the pedestrian and the sensor by emitting ultrasonic waves and detecting their reflected waves, so as to estimate the number of pedestrians. This method has good effects under specific conditions, but is greatly affected by environmental noise.

[0007] With the development of technologies such as big data and artificial intelligence, the perception of the crowd flow situation based on intelligent video analysis, these technologies not only improve the efficiency and accuracy of crowd monitoring. However, the above technologies still have the following problems when performing crowd monitoring: 1. The layout positions of monitoring devices are limited and cannot cover all areas, and it is impossible to detect the crowd flow in non-monitored areas; 2. The display method of the monitored crowd flow data is single. For example, different colors are used to mark and display each area on a two-dimensional map, and the display effect is poor. Summary of the Invention

[0008] The purpose of the present invention is to provide a safety risk monitoring and early warning method and system based on a digital earth map to solve the problems that the existing monitoring methods cannot detect the crowd flow in non-monitored areas and the display effect of crowd flow data is poor.

[0009] To achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a safety risk monitoring and early warning method based on a digital earth map, and the method includes: Obtain pedestrian video information monitored at multiple sampling points within a preset area, and determine the movement trajectories of each pedestrian based on the pedestrian video information; Integrate the movement trajectories of each pedestrian into the digital earth map of the preset area pre-constructed to obtain a trajectory digital map. Multiple map areas are divided in the digital earth map, and the trajectory digital map in each map area is visually displayed; Determine the crowding density in each map area during a future period based on the trajectory digital map; Construct the crowd safety level of each map area based on the crowding density in each map area during the future period; When the crowd safety level in any map area reaches the risk level, generate a risk early warning message and send the risk early warning message to security personnel for early warning.

[0010] Preferably, determining the movement trajectories of each pedestrian based on the pedestrian video information includes: Preprocess the pedestrian video information to obtain a pedestrian video sequence; Extract multiple target pedestrians from the pedestrian video sequence based on the Gaussian background difference algorithm; Create graph nodes for each target pedestrian, establish edges between the graph nodes and the assigned weights of each edge according to preset rules, and construct an objective function with the edges between the graph nodes and the assigned weights of each edge; Solve the objective function based on the graph optimization algorithm to obtain several tracking trajectories; Associate each target pedestrian with each tracking trajectory based on the Hungarian algorithm to obtain the movement trajectories of each pedestrian.

[0011] Preferably, the preset rules at least include: appearance similarity rule, spatial proximity rule, and motion similarity rule.

[0012] Preferably, extracting multiple target pedestrians from the pedestrian video sequence based on the Gaussian background difference algorithm includes: Extract the first k image frames of the pedestrian video sequence, input the first k image frames into a pre-constructed Gaussian background model for background update to obtain a background image; where k is a positive integer; Perform a difference operation on all image frames in the pedestrian video sequence and the background image to obtain a pedestrian moving foreground image; Perform filtering processing on the pedestrian moving foreground image to obtain a processed pedestrian moving foreground image; Based on the contour extraction and filling method, target extraction is performed on the processed foreground image of pedestrian movement to obtain multiple target pedestrians.

[0013] Preferably, the expression of the Gaussian background model is: ; In the formula, is the Gaussian background model, representing t the observation value of the pixel point of the image frame at the probability of being judged as the background, is t the weight of the n th Gaussian distribution at is t the probability density of the Gaussian distribution of the k frame image frame at N is the total number of Gaussian distributions constructed for all pixel points of each image frame; Among them, .

[0014] Preferably, the filtering process at least includes: median filtering process, morphological filtering process, and histogram redundancy processing.

[0015] Preferably, the movement trajectories of each pedestrian are integrated into the digital earth map of a preset area constructed in advance to obtain a trajectory digital map, including: Extracting the image coordinates in the movement trajectories of each pedestrian; Mapping the image coordinates in the movement trajectories of each pedestrian to the digital earth map to obtain the ground plane coordinates of each pedestrian; Determining the ground plane trajectory of each pedestrian according to the ground plane coordinates of each pedestrian; Constructing a trajectory digital map according to the ground plane trajectories of each pedestrian.

[0016] Preferably, determining the crowding density in each map area in the future period based on the trajectory digital map includes: Calculating the walking speed and walking direction of each pedestrian according to the ground plane trajectories of each pedestrian in the trajectory digital map; Dividing all pedestrians with the same walking direction in each map area into a group, and calculating the group speed of each group in each map area according to the walking speed of each pedestrian; Calculating the total number of pedestrians in each map area in the future period based on the group speed of each group in each map area; Determining the crowding density in each map area in the future period based on the total number of pedestrians in each map area in the future period and the area of each map area.

[0017] Preferably, the method further includes: determining the risk level of each map area, including: extracting the width of each map area; determining the maximum pedestrian flow of each map area based on the width of each map area and the group velocity; determining the risk level of each map area according to the maximum pedestrian flow of each map area.

[0018] In a second aspect, the present invention provides a security risk monitoring and early warning system based on a digital earth map, which is used to implement the above-mentioned security risk monitoring and early warning method based on a digital earth map. The system includes: A trajectory calculation module, configured to obtain pedestrian video information monitored at multiple sampling points within a preset area, and determine the movement trajectories of each pedestrian based on the pedestrian video information; A map construction module, configured to integrate the movement trajectories of each pedestrian into the digital earth map of the preset area pre-constructed, obtain a trajectory digital map, where multiple map areas are divided in the digital earth map, and visually display the trajectory digital map in each map area; A density calculation module, configured to determine the congestion density within each map area in a future time period based on the trajectory digital map; A level construction module, configured to construct the pedestrian flow safety level of each map area based on the congestion density within each map area in a future time period; A risk early warning module, configured to generate a risk early warning message when the pedestrian flow safety level of any map area reaches the risk level, and send the risk early warning message to security personnel for early warning.

[0019] Beneficial effects: By monitoring pedestrian video information at multiple sampling points, determining the movement trajectories of each pedestrian according to the pedestrian video information, and then integrating the movement trajectories of each pedestrian into the digital earth map to obtain a trajectory digital map, in the trajectory digital map of the digital earth map, managers can see the real-time movement status of each pedestrian, and can vividly display the pedestrian flow information, so as to solve the problem that the pedestrian flow in non-monitored areas cannot be detected; Secondly, according to the congestion density within each map area determined based on the trajectory digital map, when the pedestrian flow safety level corresponding to the congestion density within each map area reaches the risk level, an early warning can be given in time, and the risk early warning message is sent to the security personnel on the scene, and the security personnel can evacuate or limit the flow of people on the scene in time to improve the safety of the scene. Description of the drawings

[0020] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and form a part of the specification, and are used together with the following specific embodiments to explain the embodiments of the present invention, but do not constitute a limitation to the embodiments of the present invention. In the accompanying drawings: Figure 1 is a flowchart of a safety risk monitoring and early warning method based on a digital earth map provided by an embodiment of the present invention; Figure 2 is a block diagram of a safety risk monitoring and early warning system based on a digital earth map provided by an embodiment of the present invention. Specific Embodiments

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the present invention in combination with the accompanying drawings and the descriptions of the embodiments or the prior art. Obviously, the following descriptions of the structures of the accompanying drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings. It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation to the present invention.

[0022] Embodiment 1 Figure 1 is a flowchart of a safety risk monitoring and early warning method based on a digital earth map provided by an embodiment of the present invention. As Figure 1 shown, this embodiment provides a safety risk monitoring and early warning method based on a digital earth map, and the method includes: Step S10: Obtain the pedestrian video information monitored at multiple sampling points within a preset area, and determine the movement trajectories of each pedestrian based on the pedestrian video information.

[0023] In this embodiment, the preset area can be areas such as commercial streets, scenic spots, parks, shopping malls, etc. Multiple sampling points are arranged within the preset area, and a camera is installed at each sampling point to collect the pedestrian videos of each sampling point. The movement trajectories of each pedestrian can be determined based on the pedestrian videos of multiple sampling points.

[0024] Specifically, determining the movement trajectories of each pedestrian based on the pedestrian video information includes: Step S101: Preprocess the pedestrian video information to obtain a pedestrian video sequence. Among them, the preprocessing includes: video resampling and filtering. Video resampling refers to the process of changing the spatial resolution or temporal resolution of video data to remove noise or compression distortion in the video to improve the image quality, thereby improving the accuracy of movement trajectory calculation.

[0025] Step S102: Extract the pedestrian video sequence based on the Gaussian background difference algorithm to obtain multiple target pedestrians.

[0026] For step S102, extracting the pedestrian video sequence based on the Gaussian background difference algorithm to obtain multiple target pedestrians includes: Step a10: Extract the first k frame images of the pedestrian video sequence, and input the first k frame images into the pre-constructed Gaussian background model for background update to obtain a background image; where k is a positive integer; Among them, the expression of the Gaussian background model is: ; In the formula, is the Gaussian background model, representing t the observation value of the pixel point of the image frame at time the probability of being judged as the background, is t the weight of the n th Gaussian distribution at time is t the probability density of the Gaussian distribution of the k frame image at time N is the total number of Gaussian distributions constructed for all pixel points of each frame image; Among them, .

[0027] In this embodiment, the background of the Gaussian background model is initialized using the first k frame images of the pedestrian video sequence. When a new image frame is input, the background image of this image frame can be obtained.

[0028] Step a20: Perform a difference operation on all the image frames in the pedestrian video sequence and the background image to obtain a foreground image of pedestrian movement; Step a30: Perform filtering processing on the foreground image of pedestrian movement to obtain a processed foreground image of pedestrian movement; among them, the filtering processing at least includes: median filtering processing, morphological filtering processing, and histogram redundancy processing.

[0029] In this embodiment, due to the influence of factors such as cameras, local background changes, ambient light, and shadows of moving targets, there will be noise, shadows, and redundant targets in the foreground image of pedestrian movement, which in turn reduces the accuracy of pedestrian target detection.

[0030] In response to this, the present invention performs filtering processing on the foreground image of pedestrian movement through methods such as median filtering processing, morphological filtering processing, and histogram redundancy processing, which can eliminate the noise, shadows, and redundant targets existing in the foreground image of pedestrian movement to improve the accuracy of pedestrian target detection.

[0031] Specifically, first perform median filtering on the foreground image of pedestrian movement, then perform morphological filtering, and finally perform histogram redundancy processing. Filtering the foreground image of pedestrian movement in this filtering order can further improve accuracy. Among them, the median filtering process is as follows: The gray value of each pixel point in the foreground image of pedestrian movement is determined by the median value of the gray values of all pixel points within its neighborhood S. The calculation formula is as follows: ; In the formula, is the gray value of the foreground image of pedestrian movement after median filtering, is the foreground image of pedestrian movement before median filtering, is the position of the pixel point after median filtering, is the position of the pixel point before median filtering, is the median filtering function.

[0032] Using the median filtering method can effectively eliminate impulse interference and salt-and-pepper noise, and protect the edge information of the image without blurring the image.

[0033] In this embodiment, the morphological filtering mainly includes erosion operation, dilation operation, opening operation and closing operation. Among them, the erosion operation can ablate the boundary of the target, make it smaller, and eliminate the noise smaller than the structural element; the dilation operation can fuse the foreground points whose distance from the target is less than the structural element, make it larger, and fill the holes within the target; the opening operation is a process of first performing the erosion operation and then the dilation operation, which can eliminate the isolated noise outside the target and smooth the boundary of the target; the closing operation is a process of first performing the dilation operation and then the erosion operation, which can eliminate the holes within the target and smooth the boundary of the target.

[0034] In this embodiment, the histogram redundancy processing method is used to perform threshold operation on the foreground image of pedestrian movement after morphological filtering. Since the color change degree of the redundant target area that appears simultaneously with the pedestrian movement target is small, its gray value in the detected foreground image of pedestrian movement is smaller compared with the pedestrian target. Therefore, these areas with smaller gray values can be removed through the histogram threshold, and the final processed foreground image of pedestrian movement can be obtained.

[0035] Step a40: Based on the contour extraction and filling method, perform target extraction on the processed foreground image of pedestrian movement to obtain multiple target pedestrians.

[0036] In this embodiment, based on the contour extraction and filling method, the specific target extraction of the processed foreground image of pedestrian movement is as follows: First, through the closing operation of contour extraction and the setting of thresholds, smaller contour areas are removed; then, the internal areas of larger contours are filled, and then the complete target area of pedestrian movement is extracted to obtain multiple target pedestrians.

[0037] Step S103: Create graph nodes for each target pedestrian, establish edges between the graph nodes and the assigned weights of each edge according to preset rules, and construct an objective function with the edges between the graph nodes and the assigned weights of each edge; among them, the preset rules include: appearance similarity rule, spatial proximity rule, and motion similarity rule; Step S104: Solve the objective function based on the graph optimization algorithm to obtain several tracking trajectories; Step S105: Based on the Hungarian algorithm, associate each target pedestrian with each tracking trajectory to obtain the motion trajectories of each pedestrian.

[0038] Step S20: Integrate the motion trajectories of each pedestrian into the digital earth map of the preset area constructed in advance to obtain a trajectory digital map. Multiple map areas are divided in the digital earth map, and the trajectory digital map in each map area is visually displayed.

[0039] Specifically, integrating the motion trajectories of each pedestrian into the digital earth map of the preset area constructed in advance to obtain a trajectory digital map includes: Step S201: Extract the image coordinates in the motion trajectories of each pedestrian; Step S202: Map the image coordinates in the motion trajectories of each pedestrian onto the digital earth map to obtain the ground plane coordinates of each pedestrian; Step S203: Determine the ground plane trajectories of each pedestrian according to the ground plane coordinates of each pedestrian; Step S204: Construct a trajectory digital map according to the ground plane trajectories of each pedestrian.

[0040] In this embodiment, since the digital earth map of the preset area has a three-dimensional map model of the area, first map the three-dimensional coordinates of the camera into the digital earth map, and then construct a plane coordinate system according to the area where the camera is located. The z-axis of the plane coordinate system is perpendicular to the ground. Since there may be inclined ground in some areas, in order to improve the accuracy of calculating the ground plane coordinates of each pedestrian, the inclination angle of the plane coordinate system of the area needs to be obtained at this time.

[0041] First, select at least three reference points on the actual ground corresponding to the plane coordinate system. The distances from the three reference points to the camera are constant. Then, obtain the shooting angle of the camera relative to the plane coordinate system (the camera can be fixed to monitor a certain area or rotate the shooting lens to change the shooting angle to obtain a larger monitoring range). Based on the angle of the camera and the three-dimensional coordinates of the three reference points, establish a mapping relationship between the image coordinates and the ground plane coordinates on the plane coordinate system.

[0042] Next, after the mapping relationship is established, the ground plane coordinates of pedestrians can be directly obtained according to the image coordinates of pedestrians in the images captured by the camera. The multi-frame images of pedestrians can draw the ground plane trajectory of pedestrians in the plane coordinate system, and then the ground plane trajectory can be integrated into the digital earth map to obtain the trajectory digital map.

[0043] Step S30: Determine the crowding density in each map area during the future time period based on the trajectory digital map.

[0044] Specifically, determining the crowding density in each map area during the future time period based on the trajectory digital map includes: Step S301: Calculate the walking speed and walking direction of each pedestrian according to the ground plane trajectories of the pedestrians in the trajectory digital map; Step S302: Divide all pedestrians with the same walking direction in each map area into a group, and calculate the group speed of each group in each map area according to the walking speeds of the pedestrians; Step S303: Calculate the total number of pedestrians in each map area during the future time period based on the group speeds of each group in each map area; Step S304: Determine the crowding density in each map area during the future time period based on the total number of pedestrians in each map area during the future time period and the area of each map area.

[0045] In this embodiment, there are pedestrians entering the area from different directions and leaving the area in different directions in each map area. In order to dynamically reflect the total number of these pedestrians and the overall movement trend of pedestrians; in this embodiment, the walking speed of each pedestrian in the area is calculated, and then divided into different groups according to the walking direction of the pedestrians. The group speed of the group with the same walking direction is the average value of the walking speeds of all pedestrians in the group; the vector sum of the group speeds of all groups reflects the movement trend of the total number of pedestrians in this area. According to the total movement trend of all groups, the change situation of the total number of pedestrians in the next area can also be predicted, and then the crowding density in each map area during the future time period can be calculated.

[0046] When the crowd density in any map area is too high, the staff shall reasonably plan the evacuation and flow restriction of pedestrians to avoid safety accidents such as congestion and trampling.

[0047] Step S40: Based on the crowd density in each map area during the future time period, construct the crowd safety level for each map area.

[0048] In this embodiment, the method further includes: determining the risk level of each map area, including: Step b10: Extract the width of each map area; Step b20: Based on the width of each map area and the group speed, determine the maximum number of pedestrians in each map area; Step b30: According to the maximum number of pedestrians in each map area, determine the risk level of each map area.

[0049] Step S50: When the crowd safety level in any map area reaches the risk level, generate a risk warning message and send the risk warning message to the security personnel for warning.

[0050] In this embodiment, when the number of pedestrians in any map area is too large, plan the evacuation direction of pedestrians or restrict pedestrians from entering according to the crowd density of each map area; the security personnel are equipped with smart terminals, and send the pedestrian evacuation direction to the smart terminals of the corresponding security personnel to guide the security personnel to guide the pedestrians in this area to other areas and reduce the crowd density in this area.

[0051] In the present invention, by monitoring pedestrian video information at multiple sampling points, determining the movement trajectories of each pedestrian according to the pedestrian video information, and then integrating the movement trajectories of each pedestrian into the digital earth map to obtain a trajectory digital map, in the trajectory digital map of the digital earth map, the management personnel can see the real-time movement state of each pedestrian, and can vividly display the crowd information; secondly, according to the crowd density in each map area during the future time period determined based on the trajectory digital map, when the crowd safety level corresponding to the crowd density in each map area reaches the risk level, an early warning can be given in time, and the risk warning message is sent to the security personnel on the scene, and the security personnel evacuate or restrict the flow of people on the scene in time to improve the safety of the scene.

[0052] Embodiment 2 Figure 2 is a block diagram of a security risk monitoring and warning system based on a digital earth map provided by an embodiment of the present invention. As Figure 2 shown, this embodiment provides a security risk monitoring and warning system based on a digital earth map for implementing the security risk monitoring and warning method based on a digital earth map in Embodiment 1. The system includes: A trajectory calculation module, configured to obtain pedestrian video information monitored at multiple sampling points within a preset area, and determine the movement trajectories of each pedestrian based on the pedestrian video information; A map construction module, configured to integrate the movement trajectories of each pedestrian into a digital earth map of the preset area constructed in advance to obtain a trajectory digital map. Multiple map areas are divided in the digital earth map, and the trajectory digital map in each map area is visually displayed; A density calculation module, configured to determine the congestion density within each map area in a future time period based on the trajectory digital map; A level construction module, configured to construct the crowd flow safety level of each map area based on the congestion density within each map area in a future time period; A risk warning module, configured to generate a risk warning message when the crowd flow safety level of any map area reaches the risk level, and send the risk warning message to security personnel for warning.

[0053] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the security risk monitoring and warning method based on the digital earth map in Embodiment 1.

[0054] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the security risk monitoring and warning method based on the digital earth map in Embodiment 1.

[0055] In the present invention, by monitoring pedestrian video information at multiple sampling points, determining the movement trajectories of each pedestrian according to the pedestrian video information, and then integrating the movement trajectories of each pedestrian into the digital earth map to obtain a trajectory digital map, in the trajectory digital map of the digital earth map, managers can see the real-time movement status of each pedestrian, and can vividly display the crowd flow information; secondly, according to the congestion density within each map area determined based on the trajectory digital map, when the crowd flow safety level corresponding to the congestion density within each map area reaches the risk level, timely warning can be given, and the risk warning message is sent to the security personnel on site, and the security personnel can evacuate or limit the flow of people on site in a timely manner to improve the safety of the site.

[0056] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an all-hardware embodiment, an all-software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 means for implementing the functions specified in one block or multiple blocks.

[0058] The above are only the embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A security risk monitoring and early warning method based on digital earth map, characterized in that: The method comprises: Obtaining pedestrian video information monitored by multiple sampling points in a preset area, and determining the movement trajectory of each pedestrian based on the pedestrian video information; Integrate the movement trajectories of each pedestrian into a pre-built digital earth map of a preset area to obtain a trajectory digital map, wherein the digital earth map is divided into a plurality of map areas, and visualize the trajectory digital map in each map area; Determining crowd density within each map area in a future time period based on the trajectory digital map; Based on the crowd density in each map area in the future time period, the pedestrian safety level of each map area is constructed; When the pedestrian safety level in any map area reaches the risk level, risk warning information is generated and sent to security personnel for warning.

2. The security risk monitoring and early warning method based on digital earth map according to claim 1 is characterized in that: Determine the movement trajectory of each pedestrian based on the pedestrian video information, including: Preprocessing pedestrian video information to obtain pedestrian video sequences; Pedestrian video sequences are extracted based on Gaussian background difference algorithm to obtain multiple target pedestrians; Create a graph node for each target pedestrian, establish the edges between the graph nodes and the allocation weights of each edge according to the preset rules, and construct the objective function based on the edges between the graph nodes and the allocation weights of each edge; Solve the objective function based on the graph optimization algorithm and obtain several tracking trajectories; Based on the Hungarian algorithm, each target pedestrian is associated with each tracking trajectory to obtain the motion trajectory of each pedestrian.

3. The security risk monitoring and early warning method based on digital earth map according to claim 2 is characterized in that: The preset rules include at least: appearance similarity rule, spatial proximity rule and motion similarity rule.

4. The security risk monitoring and early warning method based on digital earth map according to claim 2 is characterized in that: The pedestrian video sequence is extracted based on the Gaussian background difference algorithm to obtain multiple target pedestrians, including: Extract the first k image frames of the pedestrian video sequence, input the first k image frames into the pre-built Gaussian background model to perform background update, and obtain a background image; wherein k is a positive integer; Perform differential operation on all image frames in the pedestrian video sequence and the background image to obtain the pedestrian motion foreground image; Performing filtering processing on the pedestrian motion foreground image to obtain a processed pedestrian motion foreground image; Based on the contour extraction and filling method, the processed pedestrian motion foreground image is subjected to target extraction to obtain multiple target pedestrians.

5. The security risk monitoring and early warning method based on digital earth map according to claim 4 is characterized in that: The expression of the Gaussian background model is: ； In the formula, is the Gaussian background model, indicating t The observed value of the pixel point of the image frame at time The probability of being judged as background, for t The moment n The weights of a Gaussian distribution, for t Moment k The probability density of the Gaussian distribution of each image frame, N The total number of Gaussian distributions constructed for all pixels of each image frame; in, .

6. The security risk monitoring and early warning method based on digital earth map according to claim 4 is characterized in that: The filtering process at least includes: median filtering process, morphological filtering process and histogram redundancy process.

7. The security risk monitoring and early warning method based on digital earth map according to claim 1 is characterized in that: The movement trajectory of each pedestrian is integrated into the pre-built digital earth map of the preset area to obtain a trajectory digital map, including: Extract the image coordinates of each pedestrian's motion trajectory; Map the image coordinates of each pedestrian's motion trajectory onto the digital earth map to obtain the ground plane coordinates of each pedestrian; Determine the ground plane trajectory of each pedestrian according to the ground plane coordinates of each pedestrian; A trajectory digital map is constructed based on the ground-level trajectory of each pedestrian.

8. The security risk monitoring and early warning method based on digital earth map according to claim 7 is characterized in that: Determine the crowd density in each map area in the future period based on the trajectory digital map, including: Calculate the walking speed and walking direction of each pedestrian according to the ground plane trajectory of each pedestrian in the trajectory digital map; All pedestrians with the same walking direction in each map area are divided into a group, and the group speed of each group in each map area is calculated according to the walking speed of each pedestrian; Based on the group speed of each group of people in each map area, calculate the total number of pedestrians in each map area in the future period; Based on the total number of pedestrians in each map area in the future time period and the area of ​​the region in each map area, the congestion density in each map area in the future time period is determined.

9. The security risk monitoring and early warning method based on digital earth map according to claim 8 is characterized in that: The method further includes: determining a risk level for each map area, including: Extract the width of each map region; Determine the maximum crowd flow for each map area based on its width and group speed; Determine the risk level of each map area based on the maximum flow of people in each map area.

10. A security risk monitoring and early warning system based on a digital earth map, used to implement the security risk monitoring and early warning method based on a digital earth map according to any one of claims 1 to 9, characterized in that: The system comprises: A trajectory calculation module is used to obtain pedestrian video information monitored by multiple sampling points in a preset area, and determine the movement trajectory of each pedestrian based on the pedestrian video information; A map construction module is used to integrate the movement trajectories of each pedestrian into a pre-constructed digital earth map of a preset area to obtain a trajectory digital map, wherein the digital earth map is divided into a plurality of map areas, and to visualize the trajectory digital map in each map area; a density calculation module, for determining the crowding density in each map area in a future period based on the trajectory digital map; A level construction module is used to construct a pedestrian safety level for each map area based on the crowd density in each map area in the future time period; The risk warning module is used to generate risk warning information when the pedestrian safety level in any map area reaches the risk level, and send the risk warning information to security personnel for warning.

Images