Data processing method, device, electronic device and storage medium

Abstract

This application provides a data processing method, apparatus, electronic device, and storage medium. The method includes: acquiring an event to be processed and multiple surveillance videos within a preset range of a target location in the event; analyzing each of the multiple surveillance videos to determine the historical and real-time movement trajectories of a target object in the surveillance footage; and determining one or more predicted movement trajectories of the target object based on the real-time and historical movement trajectories. This method, by analyzing multiple surveillance videos within a preset range of the target location to obtain the historical and real-time movement trajectories of the target object, and thus obtaining the most likely predicted movement trajectory of the target object, helps relevant agencies to promptly locate the target object.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to a data processing method, device, electronic device and storage medium. Background Art

[0002] As the scale of video surveillance continues to expand, it has also brought about a huge amount of video data, which makes the traditional online video search based on human-wave tactics like looking for a needle in a haystack. In related technologies, when it is necessary to lock a target object in a surveillance video, the surveillance video about the target object is manually searched and the acquisition location of the surveillance video containing the target object is further obtained, and the target object is captured through these acquisition locations.

[0003] In the above-mentioned related technologies, when locating the target object, on the one hand, it takes a lot of time to manually search for surveillance videos about the target object, and on the other hand, manpower is directly deployed at the collection points of surveillance videos where the target object has appeared, which is not only inefficient but also seriously affects the probability of capturing the target object. Therefore, how to better analyze the surveillance video data and lock the target object to meet the needs of relevant departments to handle tasks has become an urgent problem to be solved. Summary of the invention

[0004] In view of the above problems, the present application provides a data processing method, device, electronic device and storage medium that overcome the above problems or at least partially solve the above problems.

[0005] The first aspect of the present application provides a data processing method, comprising:

[0006] Acquire a pending event and a plurality of surveillance videos within a preset range from a target location in the pending event;

[0007] Analyze the multiple surveillance videos respectively to determine the historical movement trajectory and real-time movement trajectory of the target object in the surveillance screen;

[0008] One or more predicted action trajectories of the target object are determined according to the real-time action trajectory and the historical action trajectory.

[0009] A second aspect of the present application provides a data processing device, including:

[0010] An acquisition module, used to acquire an event to be processed and a plurality of surveillance videos within a preset range from a target location in the event to be processed;

[0011] An analysis module, used to analyze the plurality of surveillance videos respectively to determine the historical movement trajectory and the real-time movement trajectory of the target object in the surveillance picture;

[0012] A determination module is used to determine one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory.

[0013] The third aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executed, implements the steps in the data processing method described in the first aspect of the present application.

[0014] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the data processing method described in the first aspect of the present application are implemented.

[0015] Through the data processing method of the present application, firstly, the event to be processed and multiple surveillance videos within a preset range from the target position in the event to be processed are obtained; then the multiple surveillance videos are analyzed separately to determine the historical action trajectory and real-time action trajectory of the target object in the surveillance screen; finally, one or more predicted action trajectories of the target object are determined based on the real-time action trajectory and the historical action trajectory. This method analyzes multiple surveillance videos with the target position within a preset range to obtain the historical action trajectory and real-time action trajectory of the target object, and then obtains the most likely predicted action trajectory of the target object, so that the relevant agencies can lock the target object in time and make better subsequent decisions for the target object (such as implementing a capture strategy for the target object, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application are briefly introduced below.

[0017] Figure 1 is a flow chart of a data processing method shown in one embodiment of the present application;

[0018] Figure 2 is a schematic diagram of a data processing system shown in an embodiment of the present application;

[0019] Figure 3 is a structural block diagram of a data processing device shown in an embodiment of the present application;

[0020] Figure 4 This is a schematic diagram of a visual network shown in an embodiment of the present application;

[0021] Figure 5 It is a schematic diagram of the hardware structure of a node server shown in an embodiment of the present application;

[0022] Figure 6It is a schematic diagram of the hardware structure of an access switch shown in an embodiment of the present application;

[0023] Figure 7 It is a schematic diagram of the hardware structure of an Ethernet protocol conversion gateway shown in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] Figure 1 1 is a flow chart of a data processing method according to an embodiment of the present application. Figure 1 , the data processing method of the present application may include the following steps:

[0026] Step S11: Acquire the event to be processed and a plurality of surveillance videos within a preset range from the target location in the event to be processed.

[0027] The execution subject of the data processing method of the present application may be a data processing system. The data processing system may obtain pending events from the alarm terminal, and the pending events include: the location of the incident, the time of the incident, the case description, etc. The alarm terminal may be an Internet alarm terminal deployed on the Internet, or a visual network alarm terminal deployed on the visual network, and this embodiment does not impose any specific restrictions on this.

[0028] The target location is the location of the incident to be processed, and the preset range can be set arbitrarily according to actual needs, for example, it can be a circular range formed by taking the target location as the center and 2 kilometers as the radius. The monitoring video can be a monitoring video collected by monitoring equipment in the Internet or a monitoring video collected by monitoring equipment in the visual network, and this embodiment does not make specific restrictions on this.

[0029] In step S11, after the data processing system obtains the event to be processed, it first determines the target location of the event to be processed, and then obtains a plurality of monitoring videos within a preset range around the target location.

[0030] Step S12: Analyze the multiple surveillance videos respectively to determine the historical movement trajectory and real-time movement trajectory of the target object in the surveillance image.

[0031] In one implementation, step S12 may include:

[0032] Determine the historical action trajectory of the target object based on the shooting location and time of the surveillance video before the event to be processed;

[0033] Determine the real-time movement trajectory of the target object based on the shooting location and time of the surveillance video after the event to be processed.

[0034] In step S12, the plurality of surveillance videos may be first analyzed separately to obtain analysis results of whether the plurality of surveillance videos include images of the target object in the event to be processed, and then the historical action trajectory and real-time action trajectory of the target object may be determined according to the analysis results.

[0035] In this embodiment, the target object may be a person, an animal, a vehicle, etc., and this embodiment does not specifically limit this. After obtaining multiple surveillance videos, the multiple surveillance videos may be viewed to determine whether the target object is in the surveillance video. If the target object can be viewed in the surveillance video, it means that the target object has appeared in the preset range. If the target object cannot be viewed in the surveillance video, it means that the target object has not appeared in the preset range.

[0036] If the target object has appeared in the preset range, it means that the target object may not have left. At this time, the historical action trajectories of the target object before sending the pending event and the real-time action trajectories after sending the pending event can be determined based on the analysis results and the shooting locations corresponding to the multiple surveillance videos. In specific implementation, the target surveillance videos containing the target object can be first determined, then the shooting locations of each target surveillance video can be determined, and finally the action trajectory of the target object can be drawn based on these shooting locations.

[0037] Among them, the historical action trajectory can be drawn based on the surveillance video collected before the pending event is sent, and the real-time action trajectory can be drawn based on the surveillance video collected at the time and after the pending event is sent. When searching for surveillance videos, the time range can also be set according to actual needs. For example, after sending the pending event, the historical action trajectory can be drawn based on the surveillance video within 24 hours before the pending event is sent, and the real-time action trajectory can be drawn based on the surveillance video from the sending of the pending event to the current moment. This embodiment does not specifically limit the time range.

[0038] Furthermore, the present application can also obtain multiple historical action tracks before sending the pending event and multiple real-time action tracks after sending the pending event according to different time periods. For example, if a pending event is received at 18:00 on a certain afternoon, the division basis is every 1 hour, and the start and end time of searching the surveillance video is 6 hours before sending the pending event, then a historical action track can be drawn based on the surveillance video from 17:00 to 18:00 on that afternoon, and another historical action track can be drawn based on the surveillance video from 16:00 to 17:00 on that afternoon, and so on, until the historical action track from 12:00 to 13:00 on that afternoon is drawn, and a total of 6 historical action tracks are obtained. Similarly, for the real-time action trajectory, the division basis is every 5 minutes. The current time is 18:20, so a real-time action trajectory can be drawn based on the surveillance video from 18:15 to 18:20, and another real-time action trajectory can be drawn based on the surveillance video from 18:10 to 18:15, and so on, until the historical action trajectory from 18:00 to 18:05 in the afternoon of the same day is drawn, and a total of 4 real-time action trajectories are obtained.

[0039] Of course, the historical action trajectory or the real-time action trajectory is divided according to different time periods. Therefore, when the time period is longer, there may be only one historical action trajectory and / or only one real-time action trajectory. That is, this application does not impose any specific restrictions on the number of historical action trajectories and real-time action trajectories.

[0040] Step S13: Determine one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory.

[0041] In this embodiment, the predicted action trajectory is an action trajectory with a high probability of the target object appearing.

[0042] In specific implementation, the predicted action trajectory can be determined from multiple real-time action trajectories and multiple historical action trajectories according to a predetermined analysis rule. For example, a part of the real-time action trajectory and a part of the historical action trajectory can be used as the predicted action trajectory, or all of the real-time action trajectory and a part of the historical action trajectory can be used as the predicted action trajectory, etc. This embodiment does not specifically limit the analysis rules.

[0043] In combination with the above embodiments, in one implementation, after determining one or more predicted action trajectories of the target object, the data processing method of the present application may further include the following steps:

[0044] Implement capture strategies for target objects according to the predicted action trajectory.

[0045] Among them, the predicted action trajectory can be used to implement corresponding strategies for the target object, such as capturing the target object.

[0046] In this embodiment, the data processing method can be applied to the capture scenario of the target object. By analyzing multiple surveillance videos, a large number of manpower can be deployed on the predicted motion trajectory where the target object is most likely to appear, and a small number of manpower can be deployed on other motion trajectories. On the one hand, this improves the success rate of capturing the target object, and on the other hand, it reduces the impact on residents when a large number of manpower is deployed on any motion trajectory.

[0047] When a witness finds a target object at the scene, he can send a pending event through the alarm terminal. After receiving the pending event, the data processing system obtains the geographic location of the scene in the pending event, and then obtains multiple surveillance videos within a preset range around the scene. By analyzing these surveillance videos, multiple historical action trajectories and multiple real-time action trajectories of the target object are obtained, and then one or more predicted action trajectories that the target object is most likely to use when leaving are screened out from the multiple historical action trajectories and multiple real-time action trajectories according to the pre-given analysis rules. Therefore, relevant agencies can implement capture strategies for target objects by predicting action trajectories, such as deploying manpower on the predicted action trajectory to increase the success rate of capturing target objects.

[0048] Through the data processing method of this embodiment, firstly, the event to be processed and multiple surveillance videos within a preset range from the target position in the event to be processed are obtained; then the multiple surveillance videos are analyzed separately to determine the historical action trajectory and real-time action trajectory of the target object in the surveillance screen; finally, one or more predicted action trajectories of the target object are determined based on the real-time action trajectory and the historical action trajectory. This method analyzes multiple surveillance videos with the target position within a preset range to obtain the historical action trajectory and real-time action trajectory of the target object, and then obtains the most likely predicted action trajectory of the target object, so that the relevant agencies can lock the target object in time and make better subsequent decisions for the target object (such as implementing a capture strategy for the target object, etc.).

[0049] In combination with the above embodiments, in one implementation, the event to be processed includes features of the target object. On this basis, the above step S12 may include:

[0050] Identify whether the image features of the surveillance video after the event to be processed contain the features of the target object;

[0051] When the characteristics of the target object are included, determining that the target object is within a preset range;

[0052] When the target object's features are not included, it is confirmed that the target object is not within the preset range;

[0053] When the analysis result indicates that the target object is within the preset range, the historical action trajectories of the target object before sending the event to be processed and the real-time action trajectories after sending the event to be processed are determined according to the analysis results and the shooting positions corresponding to the multiple surveillance videos.

[0054] In this embodiment, taking the target object as a person as an example, the characteristics may be height, body shape, gender, clothing, age, appearance, etc.; taking the target object as a car as an example, the characteristics may be color, model, license plate number, etc. This embodiment does not impose specific restrictions on the characteristics of the target object.

[0055] When sending an event to be processed, the user can add the characteristics of the target object on the alarm terminal and then submit the event to be processed. In this way, the data processing system can analyze the surveillance video according to the characteristics of the target object. Among them, the alarm terminal has an image acquisition function, which can collect images of the surrounding people in real time. Therefore, when sending an event to be processed, the user can also add the image information of the target object in the event to be processed, so as to better locate the target object.

[0056] In specific implementation, the data processing system can first analyze the images of the surveillance video after the event to be processed is sent, and identify whether the image features of these surveillance videos contain the features of the target object. If they are contained, it means that the target object is still within the preset range after the event to be processed is sent and has not left. If not, it means that the target object has left the preset range after the event to be processed is sent.

[0057] In this embodiment, if the target object is still within the preset range, it is necessary to further obtain the real-time action trajectory of the target object, predict the most likely trajectory of the target object, and implement subsequent strategies.

[0058] In combination with the above embodiments, in one implementation, the event to be processed also includes the characteristics of the target object. On this basis, according to the shooting positions and times corresponding to the multiple surveillance videos before the event to be processed, the historical action trajectory of the target object can be determined, which may include the following steps:

[0059] Acquire multiple frames of historical monitoring images containing features of the target object from historical monitoring videos before the event to be processed;

[0060] According to different time periods, multiple frames of historical monitoring images are divided into multiple different image groups, and the timestamps carried by each frame of historical monitoring images in each image group are all within a time period;

[0061] For each image group, the corresponding shooting positions of multiple frames of historical monitoring images are connected into a historical action trajectory in the order of shooting time.

[0062] In combination with the above embodiments, in another implementation, the event to be processed also includes the characteristics of the target object. On this basis, according to the shooting positions and times corresponding to the multiple surveillance videos after the event to be processed, the historical action trajectory and real-time action trajectory of the target object are determined, including:

[0063] Acquire multiple frames of real-time monitoring images containing features of the target object from the real-time monitoring video after the event to be processed;

[0064] According to different time periods, multiple frames of real-time monitoring images are divided into multiple different image groups, and the timestamps carried by each frame of real-time monitoring images in each image group are all within a time period;

[0065] For each image group, the corresponding shooting positions of multiple frames of real-time monitoring images are connected into a real-time action trajectory in the order of shooting time.

[0066] The above-mentioned method of obtaining the historical action trajectory and the method of obtaining the real-time action trajectory may be applied separately or simultaneously, and this embodiment does not impose any specific limitation on this.

[0067] The real-time monitoring video includes both the monitoring video after the pending event is sent and the monitoring video at the time when the pending event is sent.

[0068] In a specific implementation, for the historical action trajectory, multiple frames of historical monitoring images containing the features of the target object can be obtained from the historical monitoring video; then, according to the preset time period, the multiple frames of historical monitoring images are divided into multiple image groups; then, for each image group, in the order of shooting time, the corresponding shooting positions of the multiple frames of historical monitoring images are connected into a historical action trajectory. In a specific implementation, for the real-time action trajectory, multiple frames of real-time monitoring images containing the features of the target object can be obtained from the real-time monitoring video; then, according to the preset time period, the multiple frames of real-time monitoring images are divided into multiple image groups; then, for each image group, in the order of shooting time, the corresponding shooting positions of the multiple frames of real-time monitoring images are connected into a real-time action trajectory.

[0069] In this embodiment, the collected multiple surveillance videos can be divided into historical surveillance videos before sending the pending event and real-time surveillance videos after sending the pending event. Each surveillance video is composed of multiple frames of surveillance images. In this embodiment, multiple frames of historical surveillance images containing features of the target object are first extracted from the historical surveillance video. Since each frame of surveillance image carries a timestamp, the multiple frames of historical surveillance images can be divided into different time periods according to the timestamps. For example, multiple frames of historical surveillance images with timestamps belonging to 10:00-11:00 are divided into one group of images, and multiple frames of historical surveillance images with timestamps belonging to 11:00-12:00 are divided into another group of images, and so on, so that multiple frames of historical surveillance images are divided into multiple image groups. Similarly, for the real-time surveillance video, multiple frames of real-time surveillance images containing the features of the target object are extracted, and the multiple frames of real-time surveillance images are divided into different time periods according to the timestamps. For example, the multiple frames of real-time surveillance images with timestamps belonging to 10:00-11:00 are divided into a group of images, and the multiple frames of real-time surveillance images with timestamps belonging to 11:00-12:00 are divided into another group of images, and so on, thereby dividing the multiple frames of real-time surveillance images into multiple image groups.

[0070] In this embodiment, after obtaining multiple image groups, for multiple frames of images in each image group, corresponding shooting positions are connected into action tracks according to the order of shooting time. Specifically, for each historical monitoring image group, the shooting positions corresponding to each of the multiple frames of historical monitoring images are connected into a historical action track according to the order of shooting time, and multiple historical action tracks are obtained. For each real-time monitoring image group, the shooting positions corresponding to each of the multiple frames of real-time monitoring images are connected into a real-time action track according to the order of shooting time, and multiple real-time action tracks are obtained.

[0071] For example, a historical monitoring image group includes image 1 (shooting time: 6:03 on X-month-X-year, shooting location: A), image 2 (shooting time: 7:21 on X-month-X-year, shooting location: B), image 3 (shooting time: 7:28 on X-month-X-year, shooting location: C), image 4 (shooting time: 7:43 on X-month-X-year, shooting location: B), and image 5 (shooting time: 6:23 on X-month-X-year, shooting location: E). Then the historical action track of the connected user between 6:00 on X-month-X-year and 8:00 on X-month-X-year is: A-E-B-C-B.

[0072] Through this embodiment, multiple historical action trajectories of the target object before sending the event to be processed and multiple real-time action trajectories of the target object after sending the event to be processed can be obtained according to the shooting time and shooting location of the monitoring image, providing technical support for subsequent data processing work.

[0073] In combination with the above embodiments, in one implementation, the present application further provides a method for obtaining a predicted action trajectory. Specifically, the above step S14 may include:

[0074] When there are multiple historical action trajectories, the confidence probability of each historical action trajectory is determined separately;

[0075] The real-time action trajectory and the historical action trajectory with a confidence probability greater than a preset threshold are determined as the predicted action trajectory.

[0076] In this embodiment, the confidence probability is used to indicate the probability of the target object appearing. The higher the confidence probability of a track, the greater the probability that the target object appears in the track, and the lower the confidence probability of a track, the smaller the probability that the target object appears in the track.

[0077] In this embodiment, a threshold can be pre-set based on experience, and all historical action trajectories with confidence probabilities greater than the preset threshold are regarded as the most likely historical action trajectories of the target object. Then, each real-time action trajectory and the most likely historical action trajectory are regarded as the most likely predicted action trajectory of the target object.

[0078] For example, in the scenario of capturing the target object, the confidence probabilities of the five historical action trajectories obtained by the data processing system are trajectory 1-50%, trajectory 2-49%, trajectory 3-68%, trajectory 4-80% and trajectory 5-89% respectively. If the preset threshold is 60%, then the trajectories greater than the preset threshold are trajectory 3-trajectory 5. In this way, in order to improve the success rate of capturing the target object, a large number of manpower can be deployed on trajectory 3-trajectory 5, and a small number of manpower can be deployed on trajectory 1-trajectory 2. In this way, the impact on residents when a large number of manpower is deployed on any trajectory can also be reduced.

[0079] In one implementation, determining the confidence probability of each historical action trajectory may include:

[0080] Count the total number of trajectory points M contained in all historical action trajectories and the total number of times N that each trajectory point appears in all historical action trajectories;

[0081] The quotient of the total number of times N that each trajectory point appears in all historical action trajectories and the total number M of trajectory points contained in all historical action trajectories is determined as the confidence probability of the trajectory point;

[0082] For each historical action trajectory, the product of the confidence probabilities of the included trajectory points is determined as the confidence probability of the historical action trajectory.

[0083] In this embodiment, when determining the confidence probability of each historical action trajectory, the total number of trajectory points in all historical action trajectories and the number of occurrences of each type of trajectory points are first counted; then the quotient of the number of occurrences of each type of trajectory points and the total number is used as the confidence probability of the type of trajectory points; then, for each historical action trajectory, the product of the confidence probabilities of the included trajectory points is used as the confidence probability of the historical action trajectory.

[0084] In this embodiment, each trajectory point is a location where the target object has appeared. The analysis of the historical action trajectory by this application can obtain the probability of the target object appearing again at each trajectory point, that is, the confidence probability. Specifically, the frequency of each trajectory point in the historical action trajectory among all the trajectory points in the historical action trajectory can be used as the probability (confidence probability) of the target object appearing again at the trajectory point.

[0085] For example, there are three historical action tracks as follows:

[0086] Historical action track 1: Location 1-Location 2-Location 3;

[0087] Historical action track 2: Location 2-Location 5-Location 4;

[0088] Historical Action Track 3: Location 3-Location 1;

[0089] Then the total number of trajectory points in the historical action trajectory is 3+3+2=8. There are 5 categories of trajectory points, namely location 1-location 5. The probability of trajectory point 1 appearing is 2 / 8=1 / 4, the probability of trajectory point 2 appearing is 2 / 8=1 / 4, the probability of trajectory point 3 appearing is 2 / 8=1 / 4, the probability of trajectory point 4 appearing is 1 / 8=1 / 8, and the probability of trajectory point 5 appearing is 1 / 8=1 / 8.

[0090] After obtaining the confidence probability of each type of trajectory point, the confidence probability of each historical action trajectory can be further obtained.

[0091] The confidence probability of historical action trajectory 1 is: confidence probability of location 1 * confidence probability of location 2 * confidence probability of location 3, that is, 1 / 4*1 / 4*1 / 4=1.56%. Similarly, the confidence probability of historical action trajectory 2 is: 1 / 4*1 / 8*1 / 8=0.39%. Similarly, the confidence probability of historical action trajectory 3 is: 1 / 4*1 / 4=6.25%.

[0092] In this embodiment, the confidence probability of the historical action trajectory can be obtained according to the historical activities of the target object within a preset range, and then the historical action trajectory of the target object most likely to appear can be obtained among all historical action trajectories. Then, strategies for the target object can be implemented on the real-time action trajectory and these most likely historical action trajectories, such as conducting key joint defense and control, capturing the target object, etc., which can effectively increase the probability of successfully capturing the target object.

[0093] In combination with the above embodiments, in one implementation, the present application further provides a method for acquiring multiple surveillance videos within a preset range. Specifically, the above step S11 may include the following steps:

[0094] Determine multiple monitoring devices within a preset range;

[0095] Retrieving multiple real-time monitoring videos of multiple monitoring devices from multiple monitoring devices;

[0096] The historical surveillance videos shot by multiple surveillance devices are obtained from a video storage server that manages multiple surveillance devices.

[0097] In this embodiment, multiple monitoring devices are deployed within the preset range, and the monitoring devices periodically send the collected monitoring videos to the video storage server for storage. Therefore, when the event to be processed is sent, the data processing system first obtains multiple real-time monitoring videos from the preset range, and then retrieves the historical monitoring videos shot by these monitoring devices from the video storage server.

[0098] Through this embodiment, multiple real-time monitoring videos and historical monitoring videos within a preset range of the target location in the event to be processed can be quickly obtained after the event to be processed is sent, thereby ensuring the smooth execution of the subsequent data processing method.

[0099] Figure 2 is a schematic diagram of a data processing system shown in an embodiment of the present application. The data processing method of the present application can also be applied to Figure 2 Data processing system in . Figure 2 The data processing system includes a joint defense and control server, a visual network alarm terminal, a monitoring access server, an intelligent analysis server, a trajectory calculation server and a confidence probability calculation server; the joint defense and control server is respectively communicated with the visual network alarm terminal, the monitoring access server, the intelligent analysis server, the trajectory calculation server and the confidence probability calculation server; the intelligent analysis server is respectively communicated with the monitoring access server and the trajectory calculation server; the trajectory calculation server is also communicated with the confidence probability calculation server.

[0100] The visual network alarm terminal is used to send pending events to the joint defense and control server;

[0101] The joint defense and control server is used to send information of pending events to the monitoring access server according to the pending events, and obtain multiple predicted action trajectories according to the results of the trajectory calculation server and the results of the confidence probability calculation server;

[0102] The monitoring access server is used to respond to the information of the event to be processed and obtain multiple monitoring videos within a preset range from the target location in the event to be processed;

[0103] The intelligent analysis server is used to analyze the multiple surveillance videos respectively to obtain analysis results of whether the multiple surveillance videos include the images of the target object in the event to be processed;

[0104] The trajectory calculation server is used to determine each historical action trajectory of the target object before sending the event to be processed, and each real-time action trajectory after sending the event to be processed, according to each analysis result and the shooting positions corresponding to the multiple surveillance videos;

[0105] The confidence probability calculation server is used to determine the confidence probability of each historical action trajectory.

[0106] Among them, the monitoring access server is communicated with the monitoring device and the video storage server respectively; the multiple monitoring videos include historical monitoring videos and real-time monitoring videos; the intelligent analysis server obtains the real-time monitoring video from the monitoring device through the monitoring access server, and obtains the historical monitoring video from the video storage server through the video storage server.

[0107] In this embodiment, a main virtual terminal and multiple non-main virtual terminals need to be configured on the monitoring access server. The main virtual terminal needs to be configured on the video storage server, intelligent analysis server, trajectory calculation server, escape probability server, and joint defense and control server. Among them, the main virtual terminal is used to manage and allocate non-main virtual terminals, as well as monitor access service visual network communication. Non-main virtual terminals are used to view each monitoring video (viewing one monitoring channel requires dynamic allocation of a non-main virtual terminal), and non-main virtual terminals are used for visual network communication.

[0108] All devices including joint defense and control servers, visual network alarm terminals, monitoring access servers, intelligent analysis servers, trajectory calculation servers, confidence probability calculation servers, monitoring equipment, video storage servers, etc. need to be authenticated by the network management server before use.

[0109] The monitoring equipment saves the collected monitoring video to the video storage server through the monitoring access server. There are many forms of visual network alarm terminals, including bayonet type, recording ceremony, etc. The bayonet type visual network alarm terminal is a terminal deployed in a place where the target object must pass through, and collects features of people passing by, such as facial capture. The recording ceremony visual network alarm terminal is an alarm terminal worn by a specific group of people, and will collect features of people in the wearer's surrounding environment, such as facial capture. When the visual network alarm terminal is working, it will constantly record information about people in the surrounding environment, such as photo information.

[0110] The following will be combined Figure 2 , taking the target object as a person as an example, a detailed description of the data processing method of this application is given.

[0111] Step 1: When a case occurs, the user can select a photo of the target object through the alarm terminal to initiate a pending event (the user can also enter the target object's features, such as height, body shape, clothing, etc. to initiate a pending event);

[0112] Step 2: The alarm terminal uploads the event to be processed to the joint defense and control server 1;

[0113] Step 3: The joint defense and control server 1 sends information about the event to be processed (including location information, photo information of the target object, etc.) to the monitoring access server 3 based on the location information in the event to be processed;

[0114] Step 4: The monitoring access server 3 obtains monitoring information data (including information configured by the monitoring device on the monitoring access server, such as IP address, port number, etc.) by parsing the information of the event to be processed, and pulls real-time monitoring video data from the monitoring device 8;

[0115] Step 5: The monitoring access server 3 sends the pulled real-time monitoring video data and the photo information of the target object obtained by parsing the information of the event to be processed to the intelligent analysis server 2;

[0116] Step 6: The intelligent analysis server 2 analyzes and calculates the real-time monitoring video data and the photo information of the target object to determine whether the real-time monitoring video data contains the target object and the acquisition location of the monitoring video data containing the target object;

[0117] Step 7: The intelligent analysis server 2 sends the analysis and calculation results to the joint defense and control server 1;

[0118] Step 8: The joint defense and control server 1 determines whether the target object has left by analyzing the calculation result data;

[0119] Step 9: If the target object does not leave, the joint defense and control server 1 will notify relevant personnel near the crime scene to handle the case;

[0120] Step 10: If the target object leaves, the joint defense and control server 1 will send the joint defense and control information to the monitoring access server 3;

[0121] Step 11: The monitoring access server 3 receives the joint defense deployment information and pulls the real-time monitoring video data of the monitoring devices near the monitoring device 8;

[0122] Step 12: The monitoring access server 3 sends the pulled real-time monitoring video data to the intelligent analysis server 2;

[0123] Step 13: The monitoring access server 3 sends monitoring history data request information to the video storage server 4 and the video storage server 5;

[0124] Step 14: After receiving the monitoring history data request information, the video storage server sends the monitoring history video data to the intelligent analysis server 2;

[0125] Step 15: The intelligent analysis server 2 analyzes and calculates the real-time monitoring video data and the monitoring historical video data, and sends the analysis and calculation results to the trajectory calculation server 1;

[0126] Step 16: The trajectory calculation server 1 calculates the real-time escape trajectory and sends it to the joint defense and control server 1;

[0127] Step 17: The trajectory calculation server 1 calculates the historical activity trajectory near the crime scene and sends it to the escape probability server 1;

[0128] Step 18: The escape probability server 1 calculates the historical activity trajectory to obtain the escape probability of the target object for each historical activity trajectory and the historical activity trajectory that the target object is most likely to use when leaving, and other escape probability result information;

[0129] Step 19: The escape probability server 1 sends the escape probability result information to the joint defense and control server 1;

[0130] Step 20: The joint defense and control server 1 displays the real-time escape route of the target object through the real-time escape trajectory;

[0131] Step 21: The joint defense and control server 1 displays the most likely escape route of the target object through the escape probability result information;

[0132] Step 22: Focus on joint defense and control based on the real-time escape route and the most likely escape route to capture the target object.

[0133] Through the above steps 1 to 22, the most likely escape routes used by the target object are obtained from all historical escape trajectories, and then key joint defense and control are carried out on the real-time escape routes and these most likely escape routes to capture the target object, which can effectively increase the probability of successfully capturing the target object.

[0134] Through this embodiment, the data processing system can not only quickly locate the target object after receiving the event to be processed, but also predict the route that the target object is most likely to use when leaving and realize intelligent joint defense and control, thereby increasing the probability of successfully capturing the target object; it also saves manpower and reduces the negative impact on society.

[0135] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0136] Based on the same inventive concept, the present application also provides a data processing device 300. Figure 3 1 is a structural block diagram of a data processing device according to an embodiment of the present application. Figure 3 , the data processing device 300 may include:

[0137] An acquisition module, used to acquire an event to be processed and a plurality of surveillance videos within a preset range from a target location in the event to be processed;

[0138] An analysis module, used to analyze the plurality of surveillance videos respectively to determine the historical movement trajectory and the real-time movement trajectory of the target object in the surveillance picture;

[0139] A determination module is used to determine one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory.

[0140] Optionally, the determining module includes:

[0141] A first determination submodule, for determining the confidence probability of each of the historical action trajectories when the historical action trajectories include multiple ones;

[0142] The second determination submodule is used to determine the real-time action trajectory and the historical action trajectory with a confidence probability greater than a preset threshold as the predicted action trajectory.

[0143] Optionally, the first determining submodule includes:

[0144] The statistics submodule is used to count the total number M of trajectory points contained in all historical action trajectories and the total number N of times each trajectory point appears in all historical action trajectories;

[0145] The third determination submodule is used to determine the quotient of the total number of times N that each trajectory point appears in all historical action trajectories and the total number M of trajectory points included in all historical action trajectories as the confidence probability of the trajectory point;

[0146] The fourth determination submodule is used to determine, for each historical action trajectory, the product of the confidence probabilities of the included trajectory points as the confidence probability of the historical action trajectory.

[0147] Optionally, the analysis module includes:

[0148] A fifth determination submodule, configured to determine the historical action trajectory of the target object according to the shooting position and time corresponding to the surveillance video before the event to be processed;

[0149] The sixth determination submodule is used to determine the real-time movement trajectory of the target object according to the shooting position and time corresponding to the surveillance video after the event to be processed.

[0150] Optionally, the event to be processed further includes the characteristics of the target object, and the fifth determining submodule includes:

[0151] A first acquisition submodule is used to acquire multiple frames of historical monitoring images containing features of the target object from the historical monitoring video before the event to be processed;

[0152] A first division submodule is used to divide the multiple frames of historical monitoring images into multiple different image groups according to different time periods, and the timestamps carried by each frame of historical monitoring images in each image group are all within a time period;

[0153] The first connection submodule is used to connect the corresponding shooting positions of multiple frames of historical monitoring images into a historical action track according to the chronological order of shooting time for each image group.

[0154] Optionally, the event to be processed further includes features of the target object, and the sixth determination submodule includes:

[0155] A second acquisition submodule is used to acquire multiple frames of real-time monitoring images containing features of the target object from the real-time monitoring video after the event to be processed;

[0156] A second division submodule is used to divide the multiple frames of real-time monitoring images into multiple different image groups according to different time periods, and the timestamps carried by each frame of real-time monitoring images in each image group are all within a time period;

[0157] The second connection submodule is used to connect the shooting positions corresponding to the multiple frames of real-time monitoring images into a real-time action trajectory according to the order of shooting time for each image group. Based on the same inventive concept, the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the data processing method described in any of the above embodiments of the present application when executing.

[0158] Based on the same inventive concept, the present application provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the data processing method described in any of the above embodiments of the present application are implemented.

[0159] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0160] The visual Internet is an important milestone in the development of the Internet. It is a real-time network that can realize real-time transmission of high-definition video and push many Internet applications towards high-definition video and high-definition face-to-face communication.

[0161] The visual network adopts real-time high-definition video exchange technology. It can integrate the required services such as high-definition video conferencing, video monitoring, intelligent monitoring analysis, emergency command, digital radio and television, delayed TV, online teaching, live broadcast, VOD on demand, TV mail, personalized recording (PVR), intranet (self-operated) channels, intelligent video broadcast control, information release and dozens of other video, voice, picture, text, communication, data and other services on a single system platform, and realize high-definition quality video playback through TV or computer.

[0162] In order to enable those skilled in the art to better understand the embodiments of the present invention, the visual network is introduced as follows:

[0163] Some of the technologies used in the visual Internet of Things are as follows:

[0164] Network Technology

[0165] The network technology innovation of visual networking improves the traditional Ethernet to cope with the huge potential video traffic on the network. Different from simple network packet switching or network circuit switching, visual networking technology uses packet switching to meet streaming needs. Visual networking technology has the flexibility, simplicity and low price of packet switching, and the quality and security guarantee of circuit switching, realizing the seamless connection of the whole network switching virtual circuit and data format.

[0166] Switching Technology

[0167] Visual networking uses the two advantages of Ethernet: asynchronism and packet switching. It eliminates Ethernet defects under the premise of full compatibility, and has seamless end-to-end connection across the entire network, directly connects to user terminals, and directly carries IP data packets. User data does not require any format conversion within the entire network. Visual networking is a more advanced form of Ethernet and a real-time switching platform that can achieve large-scale real-time transmission of high-definition video across the entire network, which is currently unattainable on the Internet, and pushes many network video applications towards high definition and unification.

[0168] Server Technology

[0169] The server technology on the visual network and unified video platform is different from the traditional server. Its streaming media transmission is based on connection-oriented. Its data processing capability is independent of traffic and communication time. A single network layer can include signaling and data transmission. For voice and video services, the complexity of streaming media processing on the visual network and unified video platform is much simpler than data processing, and its efficiency is more than 100 times higher than that of traditional servers.

[0170] Storage Technology

[0171] The ultra-high-speed storage technology of the unified video platform adopts the most advanced real-time operating system to adapt to the media content with ultra-large capacity and ultra-large flow, mapping the program information in the server command to the specific hard disk space. The media content no longer passes through the server, but is directly delivered to the user terminal in an instant. The user generally waits for less than 0.2 seconds. The optimized sector distribution greatly reduces the mechanical movement of the hard disk head seeking, and the resource consumption accounts for only 20% of the same level of IP Internet, but it generates 3 times more concurrent traffic than the traditional hard disk array, and the overall efficiency is improved by more than 10 times.

[0172] Network Security Technology

[0173] The structural design of the visual network completely eradicates the network security issues that plague the Internet through separate licensing for each service and complete isolation of equipment and user data. Antivirus programs and firewalls are generally not required, and attacks by hackers and viruses are eliminated, providing users with a structurally worry-free and secure network.

[0174] Service Innovation Technology

[0175] The unified video platform integrates services and transmission. Whether it is a single user, a private network user or a network, it is just an automatic connection. User terminals, set-top boxes or PCs are directly connected to the unified video platform to obtain rich and colorful multimedia video services in various forms. The unified video platform uses a "recipe-style" table configuration mode to replace traditional complex application programming. Complex applications can be implemented with very little code, realizing "unlimited" new business innovation.

[0176] The networking of the visual network is as follows:

[0177] Visual networking is a centrally controlled network structure. The network can be a tree network, star network, ring network, etc., but on this basis, there needs to be a centralized control node in the network to control the entire network.

[0178] Figure 4 FIG. 1 is a schematic diagram of a visual network shown in an embodiment of the present application. Figure 4 As shown in the figure, the visual network is divided into two parts: access network and metropolitan area network.

[0179] The equipment in the access network can be divided into three categories: node server, access switch, and terminal (including various set-top boxes, encoding boards, storage, etc.). The node server is connected to the access switch, and the access switch can be connected to multiple terminals and can be connected to Ethernet.

[0180] The node server is a node in the access network that performs centralized control functions and can control access switches and terminals. The node server can be directly connected to the access switch or directly connected to the terminal.

[0181] Similarly, the devices in the metropolitan area network can also be divided into three categories: metropolitan area servers, node switches, and node servers. The metropolitan area servers are connected to the node switches, and the node switches can be connected to multiple node servers.

[0182] The node server is a node server of the access network part, that is, the node server belongs to both the access network part and the metropolitan area network part.

[0183] The metropolitan area server is a node with centralized control function in the metropolitan area network, which can control the node switch and node server. The metropolitan area server can be directly connected to the node switch or the node server.

[0184] It can be seen from this that the entire visual network is a hierarchical and centrally controlled network structure, and the network controlled by the node server and the metropolitan area server can be various structures such as tree, star, and ring.

[0185] Figuratively speaking, the access network part can form a unified video platform (the part in the dotted circle), and multiple unified video platforms can form a visual network; each unified video platform can be interconnected through the metropolitan area and wide area visual networks.

[0186] Visual networking equipment classification

[0187] 1.1 The devices in the visual network of the embodiment of the present invention can be mainly divided into three categories: servers, switches (including Ethernet gateways), terminals (including various set-top boxes, encoding boards, storage devices, etc.). The visual network as a whole can be divided into a metropolitan area network (or a national network, a global network, etc.) and an access network.

[0188] 1.2 The equipment in the access network can be mainly divided into three categories: node server, access switch (including Ethernet gateway), terminal (including various set-top boxes, encoding boards, storage, etc.).

[0189] The specific hardware structure of each access network device is as follows:

[0190] Node server:

[0191] Figure 5 FIG. 1 is a schematic diagram of the hardware structure of a node server shown in an embodiment of the present application. Figure 5 As shown, it mainly includes a network interface module 501, a switching engine module 502, a CPU module 503, and a disk array module 504;

[0192] Among them, the packets coming from the network interface module 501, the CPU module 503, and the disk array module 504 all enter the switching engine module 502; the switching engine module 502 performs an operation of checking the address table 505 on the incoming packets, thereby obtaining the packet's guidance information; and according to the packet's guidance information, the packet is stored in the queue of the corresponding packet buffer 506; if the queue of the packet buffer 506 is close to full, it is discarded; the switching engine module 502 polls all packet buffer queues, and forwards if the following conditions are met: 1) the port sending buffer is not full; 2) the queue packet counter is greater than zero. The disk array module 504 mainly implements the control of the hard disk, including the initialization, reading and writing of the hard disk; the CPU module 503 is mainly responsible for the protocol processing between the access switch and the terminal (not shown in the figure), the configuration of the address table 505 (including the downstream protocol packet address table, the upstream protocol packet address table, and the data packet address table), and the configuration of the disk array module 504.

[0193] Access switch:

[0194] Figure 6 FIG. 1 is a schematic diagram of the hardware structure of an access switch shown in an embodiment of the present application. Figure 6 As shown, it mainly includes a network interface module (a downlink network interface module 601, an uplink network interface module 602), a switching engine module 603 and a CPU module 604;

[0195] Among them, the packet (uplink data) coming from the downlink network interface module 601 enters the packet detection module 605; the packet detection module 605 detects whether the destination address (DA), source address (SA), data packet type and packet length of the packet meet the requirements. If they meet the requirements, the corresponding stream identifier (stream-id) is allocated and enters the switching engine module 603, otherwise it is discarded; the packet (downlink data) coming from the uplink network interface module 602 enters the switching engine module 603; the data packet coming from the CPU module 604 enters the switching engine module 603; the switching engine module 603 processes the incoming packet The address table 606 is looked up to obtain the packet's guidance information; if the packet entering the switching engine module 603 is from the downstream network interface to the upstream network interface, the packet is stored in the queue of the corresponding packet buffer 607 in combination with the stream identifier (stream-id); if the queue of the packet buffer 607 is nearly full, it is discarded; if the packet entering the switching engine module 603 is not from the downstream network interface to the upstream network interface, the data packet is stored in the queue of the corresponding packet buffer 607 according to the packet's guidance information; if the queue of the packet buffer 607 is nearly full, it is discarded.

[0196] The switching engine module 603 polls all packet buffer queues. In the embodiment of the present invention, there are two situations:

[0197] If the queue is from the downlink network interface to the uplink network interface, it is forwarded if the following conditions are met: 1) the port send buffer is not full; 2) the queue packet counter is greater than zero; 3) a token generated by the bit rate control module is obtained;

[0198] If the queue is not going from the downstream network interface to the upstream network interface, it is forwarded if the following conditions are met: 1) the port send buffer is not full; 2) the queue packet counter is greater than zero.

[0199] The code rate control module 608 is configured by the CPU module 604 to generate tokens for all packet buffer queues from the downstream network interface to the upstream network interface within a programmable interval to control the code rate of the upstream forwarding.

[0200] The CPU module 604 is mainly responsible for protocol processing with the node server, configuration of the address table 606 , and configuration of the bit rate control module 608 .

[0201] Ethernet protocol conversion gateway :

[0202] Figure 7 FIG. 1 is a schematic diagram of the hardware structure of an Ethernet protocol conversion gateway shown in an embodiment of the present application. Figure 7 As shown, it mainly includes a network interface module (a downlink network interface module 701, an uplink network interface module 702), a switching engine module 703, a CPU module 704, a packet detection module 705, a bit rate control module 708, an address table 706, a packet buffer 707, a MAC adding module 709, and a MAC deleting module 710.

[0203] The data packet from the downlink network interface module 701 enters the packet detection module 705; the packet detection module 705 detects whether the Ethernet MAC DA, Ethernet MAC SA, Ethernet length or frame type, visual network destination address DA, visual network source address SA, visual network data packet type and packet length of the data packet meet the requirements, and if they meet the requirements, the corresponding stream identifier (stream-id) is assigned; then, the MAC deletion module 710 subtracts the MAC DA, MAC SA, length or frame type (2 bytes) and enters the corresponding receiving buffer, otherwise it is discarded;

[0204] The downlink network interface module 701 detects the sending buffer of the port. If there is a packet, it obtains the Ethernet MAC DA of the corresponding terminal based on the visual network destination address DA of the packet, adds the Ethernet MAC DA of the terminal, the MACSA of the Ethernet protocol conversion gateway, and the Ethernet length or frame type, and sends it.

[0205] The functions of other modules in the Ethernet protocol conversion gateway are similar to those of the access switch.

[0206] terminal:

[0207] It mainly includes network interface module, business processing module and CPU module; for example, the set-top box mainly includes network interface module, audio and video encoding and decoding engine module, CPU module; the encoding board mainly includes network interface module, audio and video encoding engine module, CPU module; the memory mainly includes network interface module, CPU module and disk array module.

[0208] 1.3 The equipment in the metropolitan area network can be divided into two categories: node server, node switch, and metropolitan area server. Among them, the node switch mainly includes network interface module, switching engine module and CPU module; the metropolitan area server mainly includes network interface module, switching engine module and CPU module.

[0209] 2. Definition of Visual Network Data Packet

[0210] 2.1 Access Network Data Packet Definition

[0211] The data packet of the access network mainly includes the following parts: destination address (DA), source address (SA), reserved bytes, payload (PDU), and CRC.

[0212] As shown in the following table, the data packet of the access network mainly includes the following parts:

[0213] DA SA Reserved Payload CRC

[0214] in:

[0215] The destination address (DA) consists of 8 bytes. The first byte indicates the type of data packet (such as various protocol packets, multicast data packets, unicast data packets, etc.), with a maximum of 256 possibilities. The second to sixth bytes are the metropolitan area network address, and the seventh and eighth bytes are the access network address.

[0216] The source address (SA) is also composed of 8 bytes and has the same definition as the destination address (DA).

[0217] The reserved bytes consist of 2 bytes;

[0218] The payload part has different lengths depending on the type of datagram. If it is a protocol packet, it is 64 bytes. If it is a single multicast data packet, it is 32+1024=1056 bytes. Of course, it is not limited to the above two types.

[0219] CRC consists of 4 bytes and its calculation method follows the standard Ethernet CRC algorithm.

[0220] 2.2 Definition of Metropolitan Area Network Data Packet

[0221] The topology of the metropolitan area network is a graph. There may be two or more connections between two devices, that is, there may be more than two connections between a node switch and a node server, a node switch and a node switch, and a node switch and a node server. However, the metropolitan area network address of a metropolitan area network device is unique. In order to accurately describe the connection relationship between metropolitan area network devices, a parameter: a label is introduced in the embodiment of the present invention to uniquely describe a metropolitan area network device.

[0222] The definition of labels in this manual is similar to the definition of labels in MPLS (Multi-Protocol Label Switch). Assuming that there are two connections between device A and device B, there are two labels for the data packet from device A to device B, and there are two labels for the data packet from device B to device A. Labels are divided into input labels and output labels. Assuming that the label of the data packet entering device A (input label) is 0x0000, the label of the data packet when leaving device A (output label) may become 0x0001. The access process of the metropolitan area network is a centralized access process, which means that the address allocation and label allocation of the metropolitan area network are dominated by the metropolitan area server, and the node switch and node server are only passively executed. This is different from the label allocation of MPLS, which is the result of mutual negotiation between the switch and the server.

[0223] As shown in the following table, the data packets of the metropolitan area network mainly include the following parts:

[0224] DA SA Reserved Label Payload CRC

[0225] That is, destination address (DA), source address (SA), reserved byte (Reserved), label, payload (PDU), CRC. The label format can refer to the following definition: the label is 32 bits, of which the upper 16 bits are reserved and only the lower 16 bits are used. It is located between the reserved bytes and payload of the data packet.

[0226] Based on the above characteristics of the visual network, one of the core concepts of an embodiment of the present invention is proposed. In accordance with the protocol of the visual network, a plurality of surveillance videos of an event to be processed and a target position in the event to be processed within a preset range are obtained; the plurality of surveillance videos are analyzed separately to determine the historical action trajectory and the real-time action trajectory of the target object in the surveillance screen; one or more predicted action trajectories of the target object are determined based on the real-time action trajectory and the historical action trajectory. The method obtains the historical action trajectory and the real-time action trajectory of the target object by analyzing a plurality of surveillance videos of the target position within a preset range, and then obtains the most likely predicted action trajectory of the target object.

[0227] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0228] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may 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 codes.

[0229] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0230] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0231] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0232] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0233] The data processing method, device, electronic device and storage medium provided by the present invention are introduced in detail above. The principle and implementation mode of the present invention are explained in detail using specific examples. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation mode and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A data processing method, characterized in that: include: Acquire a pending event and a plurality of surveillance videos within a preset range from a target location in the pending event; The event to be processed is submitted by the user after adding the characteristics of the target object on the alarm terminal; Analyze the multiple surveillance videos respectively to determine the historical movement trajectory and real-time movement trajectory of the target object in the surveillance screen; Determining one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory; The step of determining one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory comprises: when the historical action trajectory includes multiple ones, determining the confidence probability of each of the historical action trajectories respectively; determining the real-time action trajectory and each of the historical action trajectories whose confidence probability is greater than a preset threshold as the predicted action trajectory; Determining the confidence probability of each of the historical action trajectories, including: counting the total number M of trajectory points included in all historical action trajectories and the total number N of times each trajectory point appears in all historical action trajectories; determining the quotient of the total number N of times each trajectory point appears in all historical action trajectories and the total number M of trajectory points included in all historical action trajectories as the confidence probability of the trajectory point; the confidence probability is used to represent the probability of the target object appearing; for each historical action trajectory, determining the product of the confidence probabilities of the included trajectory points as the confidence probability of the historical action trajectory; The multiple surveillance videos are analyzed separately to determine the historical movement trajectory and real-time movement trajectory of the target object in the surveillance screen, including: determining the historical movement trajectory of the target object according to the shooting position and time corresponding to the surveillance video before the event to be processed; determining the real-time movement trajectory of the target object according to the shooting position and time corresponding to the surveillance video after the event to be processed.

2. The method according to claim 1, characterized in that The event to be processed also includes the characteristics of the target object, and the historical action trajectory of the target object is determined according to the shooting positions and times corresponding to the multiple surveillance videos before the event to be processed, including: Acquire multiple frames of historical monitoring images containing features of the target object from historical monitoring videos before the event to be processed; According to different time periods, the multiple frames of historical monitoring images are divided into a plurality of different image groups, and the timestamps carried by each frame of historical monitoring images in each image group are all within a time period; For each image group, the corresponding shooting positions of multiple frames of historical monitoring images are connected into a historical action trajectory in the order of shooting time.

3. The method according to claim 1 or 2, characterized in that: The event to be processed also includes the characteristics of the target object, and the historical action trajectory and real-time action trajectory of the target object are determined according to the shooting positions and times corresponding to the multiple surveillance videos after the event to be processed, including: Acquire multiple frames of real-time monitoring images containing features of the target object from the real-time monitoring video after the event to be processed; According to different time periods, the multiple frames of real-time monitoring images are divided into a plurality of different image groups, and the timestamps carried by each frame of real-time monitoring images in each image group are all within a time period; For each image group, the corresponding shooting positions of multiple frames of real-time monitoring images are connected into a real-time action trajectory in the order of shooting time.

4. A data processing device, characterized in that: include: An acquisition module, used to acquire an event to be processed and a plurality of surveillance videos within a preset range from a target location in the event to be processed; The event to be processed is submitted by the user after adding the characteristics of the target object on the alarm terminal; An analysis module, used to analyze the plurality of surveillance videos respectively to determine the historical movement trajectory and the real-time movement trajectory of the target object in the surveillance picture; A determination module, configured to determine one or more predicted action trajectories of the target object according to the real-time action trajectory and the historical action trajectory; The determination module comprises: A first determination submodule, for determining the confidence probability of each of the historical action trajectories when the historical action trajectories include multiple ones; the confidence probability is used to represent the probability of the target object appearing; A second determination submodule, configured to determine the real-time action trajectory and each historical action trajectory with a confidence probability greater than a preset threshold as the predicted action trajectory; The first determining submodule includes: The statistics submodule is used to count the total number M of trajectory points contained in all historical action trajectories and the total number N of times each trajectory point appears in all historical action trajectories; The third determination submodule is used to determine the quotient of the total number of times N that each trajectory point appears in all historical action trajectories and the total number M of trajectory points included in all historical action trajectories as the confidence probability of the trajectory point; A fourth determination submodule is used to determine, for each historical action trajectory, the product of the confidence probabilities of the included trajectory points as the confidence probability of the historical action trajectory; The analysis module comprises: A fifth determination submodule, configured to determine the historical action trajectory of the target object according to the shooting position and time corresponding to the surveillance video before the event to be processed; The sixth determination submodule is used to determine the real-time movement trajectory of the target object according to the shooting position and time corresponding to the surveillance video after the event to be processed.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the data processing method as described in any one of claims 1 to 3 are implemented.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When executed, the processor implements the steps in the data processing method according to any one of claims 1 to 3.

Images