Video tracking method, device, equipment, machine-readable storage medium and system

Abstract

The present application provides a video tracking method, apparatus, device, machine-readable storage medium, and system. The method includes: receiving a video patrol task, the video patrol task including image information of a target to be tracked; upon detecting a moving target within the patrol range, determining the distance between the moving target and the device; and, if the distance between the moving target and the device exceeds the effective recognition range of the device, blurring the details of the image of the target to be tracked, matching the moving target based on the blurred image, and tracking the moving target if a match is successful. This method can expand the tracking range of video tracking and improve the reliability of video tracking.

Description

Technical Field

[0001] The present application relates to the field of video surveillance technology, and in particular to a video tracking method, apparatus, device, machine-readable storage medium, and system. Background Art

[0002] With the rapid development of technologies such as network technology and image processing technology, video surveillance has gradually become popular, the coverage of video surveillance has gradually increased, and the role of video surveillance in urban security has become increasingly important. Based on the video data of the deployed video surveillance equipment, the tracking of designated targets can be achieved. Summary of the Invention

[0003] In view of this, the present application provides a video tracking method, apparatus, device, machine-readable storage medium and system.

[0004] According to a first aspect of an embodiment of the present application, a video tracking method is provided, which is applied to monitoring a front-end device. The method includes:

[0005] receiving a video inspection task, wherein the video inspection task includes image information of a target to be tracked;

[0006] When a moving target is detected within the inspection range, determining the distance between the moving target and the device;

[0007] When the distance between the mobile target and the device exceeds the effective recognition range of the device, the details of the image of the target to be tracked are blurred, the mobile target is matched based on the blurred image, and if the match is successful, the mobile target is tracked.

[0008] According to a second aspect of an embodiment of the present application, a video tracking device is provided, which is deployed in a monitoring front-end device, and the device includes:

[0009] A receiving unit, configured to receive a video inspection task, wherein the video inspection task includes image information of a target to be tracked;

[0010] A determination unit, configured to determine the distance between the moving target and the monitoring front-end device when a moving target is detected to be present within the inspection range;

[0011] a matching unit configured to, when the distance between the mobile target and the monitoring front-end device exceeds the effective recognition range of the monitoring front-end device, perform detail blurring processing on the image of the target to be tracked, and match the mobile target based on the blurred image;

[0012] The tracking unit is used to track the moving target when the matching unit successfully matches.

[0013] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a processor and a memory, wherein the memory stores machine-executable instructions that can be executed by the processor, and the processor is used to execute the machine-executable instructions to implement the above-mentioned video tracking method.

[0014] According to a fourth aspect of an embodiment of the present application, a machine-readable storage medium is provided, wherein the machine-readable storage medium stores machine-executable instructions, and the machine-executable instructions implement the above-mentioned video tracking method when executed by a processor.

[0015] According to a fifth aspect of an embodiment of the present application, a video tracking system is provided, comprising: a monitoring front-end device and a central server; wherein:

[0016] The central server is used to send video inspection tasks to the monitoring front-end device, wherein the video inspection tasks include image information of the target to be tracked;

[0017] The monitoring front-end device is used to determine the distance between the moving target and the device when a moving target is detected to appear within the patrol range; when the distance between the moving target and the device exceeds the effective recognition range of the device, blur the details of the image of the target to be tracked, match the moving target based on the blurred image, and track the moving target if the match is successful.

[0018] The video tracking method of the embodiment of the present application, when detecting that a moving target appears within the patrol range and the distance between the moving target and the device exceeds the effective recognition range of the device, blurs the details of the image of the target to be tracked, matches the moving target based on the blurred image, and tracks the moving target if the match is successful, thereby expanding the tracking range of video tracking and improving the reliability of video tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of a video tracking method provided in an embodiment of the present application;

[0020] Figure 2 is a structural diagram of a video tracking device provided in an embodiment of the present application;

[0021] Figure 3 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application;

[0022] Figure 4 It is a structural diagram of a video tracking system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0024] The terms used in this application are for the purpose of describing particular embodiments only and are not intended to limit this application. As used in this application and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0025] In order to enable those skilled in the art to better understand the technical solutions provided by the embodiments of the present application, and to make the above-mentioned purposes, features and advantages of the embodiments of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application are further described in detail below with reference to the accompanying drawings.

[0026] See Figure 1 , is a flow chart of a video tracking method provided in an embodiment of the present application, wherein the video tracking method can be applied to monitoring front-end devices, such as Figure 1 As shown, the video tracking method may include the following steps:

[0027] Step S100: Receive a video inspection task, which includes image information of a target to be tracked.

[0028] In an embodiment of the present application, when it is necessary to track a specified target (referred to as the target to be tracked in this article), a video inspection task including image information of the target to be tracked can be sent to the monitoring front-end device, so that the monitoring front-end device can analyze the video data based on the image information of the target to be tracked to achieve detection and tracking of the target to be tracked.

[0029] For example, the video inspection task can be sent to the monitoring front-end device through the central server.

[0030] For example, relevant personnel can configure the parameters of the video patrol task through the video patrol task configuration interface provided by the central server, and send the video patrol task to the monitoring front-end device through the central server.

[0031] Exemplarily, the image information of the target to be tracked included in the video inspection task may be a picture of the target to be tracked or a storage address of the picture of the target to be tracked, etc.

[0032] For example, the target to be tracked may include but is not limited to people, vehicles, or animals.

[0033] Exemplarily, the central server may include a server other than the multiple monitoring front-end devices used to perform video tracking tasks (such as a background server, also referred to as a central platform) or one of the multiple monitoring front-end devices used to perform video tracking tasks (also referred to as a judgment server).

[0034] For example, multiple monitoring front-end devices used to perform video tracking tasks can elect a decision server according to a preset election strategy, such as electing the monitoring front-end device with the smallest average distance to other monitoring front-end devices as the decision server.

[0035] It should be noted that for the scenario where the central server is a server other than the multiple monitoring front-end devices used to perform video tracking tasks, each monitoring front-end device needs to establish a network connection with the central server, and the central server can save the static parameters of each monitoring front-end device, such as GPS (Global Positioning System) information, altitude, horizontal distance between monitoring front-end devices and other information; for the scenario where the central server is one of the multiple monitoring front-end devices used to perform video tracking tasks, each monitoring front-end device can establish a network connection with each other and save the static parameters of each monitoring front-end device.

[0036] Step S110: When a moving target is detected to be within the inspection range, the distance between the moving target and the device is determined.

[0037] Step S120: When the distance between the moving target and the device exceeds the effective recognition range of the device, the image of the target to be tracked is blurred in detail, the moving target is matched based on the blurred image, and the moving target is tracked if the match is successful.

[0038] In the embodiment of the present application, the effective recognition range refers to the distance at which a clear target can be detected.

[0039] For example, taking the license plate as an example, when the vehicle is far away from the monitoring front-end device and the license plate characters cannot be determined, the license plate will not be recognized. At this time, it can be considered that the distance between the target and the monitoring front-end device exceeds the effective recognition range of the monitoring front-end device.

[0040] In the embodiments of the present application, it is taken into account that in traditional video tracking solutions, it is usually necessary to identify and track the moving target only when the moving target enters the effective identification range of the monitoring front-end device (that is, the distance between the moving target and the monitoring front-end device does not exceed the effective identification range of the monitoring front-end device). The tracking range is relatively limited and can be easily circumvented.

[0041] Therefore, in order to expand the tracking range of video tracking and improve the reliability of video tracking, in an embodiment of the present application, for mobile targets that enter the inspection range of the monitoring front-end device but do not enter the effective identification range of the monitoring front-end device, the mobile targets can be identified and tracked through fuzzy matching.

[0042] Accordingly, when the monitoring front-end device detects that a moving target appears within the inspection range, the distance between the moving target and the device can be determined.

[0043] Taking into account that the target is far away from the monitoring front-end device, its image features will also change to a certain extent. Therefore, in order to improve the accuracy of target recognition, when the monitoring front-end device determines that the distance between the moving target and the device exceeds the effective recognition range of the device, the image of the target to be tracked can be blurred in detail, and the moving target that enters the inspection range can be matched based on the blurred image. If the match is successful, the moving target can be tracked.

[0044] In one example, the detail blurring process for the image of the target to be tracked may include:

[0045] Reduce the pixel size of the image of the target to be tracked;

[0046] And / or, a blur algorithm is used to blur details of the image of the target to be tracked.

[0047] Exemplarily, the details of the image of the target to be tracked can be blurred by reducing the pixels of the image of the target to be tracked, and / or using a fuzzy algorithm (such as a Gaussian algorithm) to blur the details of the image of the target to be tracked. Then, the above-mentioned moving targets entering the patrol range can be matched based on the blurred image to improve the accuracy of identifying targets outside the effective identification range.

[0048] It should be noted that in an embodiment of the present application, when the monitoring front-end device determines that the distance between the mobile target and the device does not exceed the effective recognition range of the device, the mobile target can be matched based on the picture of the target to be tracked (which can be called exact matching), and if the match is successful, the mobile target can be tracked.

[0049] It can be seen that in Figure 1In the method flow shown, when a moving target is detected to appear within the patrol range and the distance between the moving target and the device exceeds the effective recognition range of the device, the image of the target to be tracked is blurred in detail, and the moving target is matched based on the blurred image. If the match is successful, the moving target is tracked, thereby expanding the tracking range of video tracking and improving the reliability of video tracking.

[0050] In some embodiments, the detail blurring process of the target image to be tracked may include:

[0051] According to the distance between the moving target and the device, the blur level matching the distance is determined;

[0052] According to the blur level, the details of the target image to be tracked are blurred.

[0053] For example, in order to further improve the accuracy of long-distance target recognition, for targets outside the effective recognition range of the monitoring front-end device, fuzzy level classification can be performed based on the distance between the target and the monitoring front-end device.

[0054] For example, for a mobile target that is within the inspection range of the monitoring front-end device and beyond the effective identification range, the monitoring front-end device can determine the distance between the mobile target and the device, and then determine the blur level that matches the distance based on the distance between the mobile target and the device, and perform detail blurring processing on the image of the target to be tracked based on the blur level.

[0055] For example, the blur level may be positively correlated with the distance between the moving target and the monitoring front-end device, and the higher the blur level, the higher the blurriness of the image obtained after the details of the image of the target to be tracked are blurred (ie, the blurred image).

[0056] In some embodiments, matching the moving target based on the blurred image may include:

[0057] Extracting salient features from the blurred image to obtain salient features of the target to be tracked; and extracting salient features from the moving target to obtain salient features of the moving target;

[0058] Similarity matching is performed between the salient features of the target to be tracked and the salient features of the moving target.

[0059] For example, considering that the mobile target is outside the effective recognition range of the monitoring front-end device, the monitoring front-end device is usually unable to accurately identify the detailed features of the mobile target, such as license plates and headlights. Therefore, in order to improve the accuracy of target recognition, for mobile targets outside the effective recognition range of the monitoring front-end device, the target can be identified based on the target's significant features, such as outline, color, etc.

[0060] Accordingly, when a blurred image is obtained in the manner described in the above embodiment, on the one hand, significant features of the blurred image can be extracted to obtain significant features of the target to be tracked; on the other hand, significant features of the moving target can be extracted based on the video data of the moving target to obtain significant features of the moving target, and similarity matching can be performed between the significant features of the target to be tracked and the significant features of the moving target.

[0061] For example, the similarity between each salient feature of the target to be tracked and each salient feature of the moving target can be determined separately. Then, based on the similarity between each salient feature of the target to be tracked and each salient feature of the moving target, a weighted similarity (also called comprehensive similarity) between the salient features of the target to be tracked and the salient features of the moving target can be determined by weighted summation.

[0062] Exemplarily, when the comprehensive similarity between the salient features of the moving target and the salient features of the target to be tracked exceeds a set similarity threshold (which may be referred to as the first similarity threshold in this article), the match can be determined to be successful; otherwise, the match can be determined to be unsuccessful.

[0063] It should be noted that for multi-target tracking scenarios, if multiple targets to be tracked appear within the inspection range of a single monitoring front-end device, the target to be tracked with the highest comprehensive similarity can be selected for tracking.

[0064] Furthermore, for mobile targets that enter the effective recognition range of the monitoring front-end device, detailed features of the mobile target can be extracted based on the mobile target's video data during identification. Furthermore, detailed features of the target to be tracked, such as animal morphology, license plates, and headlights, can be extracted based on an image of the target to be tracked. Similarity matching is then performed between the detailed features of the mobile target and the detailed features of the target to be tracked. If the combined similarity between the detailed features of the mobile target and the detailed features of the target to be tracked exceeds a set similarity threshold (herein referred to as the second similarity threshold), the match can be determined to be successful; otherwise, the match can be determined to be unsuccessful.

[0065] In one example, the video inspection task may further include designated significant features that require fuzzy matching;

[0066] The above-mentioned extraction of salient features from the blurred image to obtain salient features of the target to be tracked; and the extraction of salient features from the moving target to obtain salient features of the moving target may include:

[0067] The designated salient features of the blurred image are extracted to obtain the designated salient features of the target to be tracked; and the designated salient features of the moving target are extracted to obtain the designated salient features of the moving target.

[0068] For example, in order to improve the flexibility and controllability of fuzzy matching, when configuring the parameters of the video patrol task, you can also select the significant features (which can be called designated significant features) used for fuzzy matching, that is, relevant personnel can select the designated significant features that need to be used in the actual matching process from the significant features supported by the system.

[0069] Accordingly, when the monitoring front-end device determines in the above manner that the distance between the moving target and the device exceeds the effective recognition range of the device, and obtains the blurred image in the above manner, it can extract the designated significant features of the blurred image based on the designated significant features included in the video patrol task to obtain the designated significant features of the target to be tracked; and extract the designated significant features of the moving target to obtain the designated significant features of the moving target, and perform similarity matching between the designated significant features of the target to be tracked and the designated significant features of the moving target.

[0070] It should be noted that, in the embodiment of the present application, if the video inspection task does not include designated significant features, it can be confirmed that all significant features supported by the system are designated significant features, that is, all significant features supported by the system need to participate in fuzzy matching.

[0071] In some embodiments, the video inspection task may further include an inspection mode, which may include a long-range inspection or a close-range inspection. Different inspection modes correspond to different effective recognition ranges.

[0072] For example, for monitoring front-end devices that support long-view mode and close-up mode, when issuing video patrol tasks, you can also specify the patrol mode of the monitoring front-end device, that is, instruct the monitoring front-end device to use long-view mode for video patrol (i.e., long-view patrol) or use close-up mode for video patrol.

[0073] Since the shooting parameters of the monitoring front-end equipment in different modes are different, the image data of the target at the same distance in different modes will also be different. Therefore, there will be different effective recognition ranges for different patrol modes.

[0074] Accordingly, for a mobile target that appears within the inspection range of the monitoring front-end device, it can be determined whether the distance exceeds the effective recognition range of the monitoring front-end device based on the inspection mode and the distance between the mobile target and the monitoring front-end device.

[0075] Furthermore, in an embodiment of the present application, the video patrol task may also include a patrol time, so that the monitoring front-end device that receives the video patrol task can perform a video patrol in the manner described in the above embodiment at the corresponding time based on the patrol time included in the video patrol task to reduce the workload of the monitoring front-end device.

[0076] In addition, for monitoring front-end devices that support steering, such as pan-tilt cameras, the video patrol task can also include a patrol direction. Thus, the monitoring front-end device that receives the video patrol task can control the pan-tilt camera to rotate to the corresponding direction based on the patrol direction included in the video patrol task, and perform video patrol in the manner described in the above embodiment.

[0077] In some embodiments, when the moving target is matched based on the blurred image and the match is successful, the following steps may also be performed:

[0078] Report the target information of the mobile target to the central server so that the central server can predict the route of the target to be tracked based on the received target information, and patrol and dispatch the monitoring front-end equipment associated with the route of the target to be tracked based on the route of the target to be tracked.

[0079] For example, when the monitoring front-end device determines that the blurred image successfully matches the moving target in the above manner, the target information of the moving target can be reported to the central server.

[0080] For example, the target information may include, but is not limited to, part or all of the information such as the detection time, the location information of the moving target, the altitude information, the moving speed, and the similarity of various significant features.

[0081] It should be noted that for mobile targets that enter the effective identification range of the monitoring front-end device, the monitoring front-end device can also report the target information of the mobile target to the central server if it successfully performs an accurate match based on the picture of the target to be tracked. At this time, the target information can include the similarity of each detailed feature.

[0082] For example, the central server can predict the route of the target to be tracked based on the target information reported by each monitoring front-end device, and based on the route of the target to be tracked, perform patrol scheduling for the monitoring front-end devices associated with the route, that is, the monitoring front-end devices deployed on the route and capable of performing patrol tasks. For example, the relevant monitoring front-end devices are pre-scheduled to enter the patrol state and wait for the target to be tracked to enter the field of view.

[0083] In one example, the central server may predict candidate travel routes, ie, possible travel routes, of the target to be tracked based on the received target information, and determine the candidate travel route with the highest credibility as the travel route of the target to be tracked.

[0084] In one example, when the moving target is matched based on the blurred image and the match is successful, the following steps may also be included:

[0085] Report the picture of the mobile target to the central server so that the central server can determine that the attributes of the target to be tracked have changed based on the picture of the mobile target, update the picture of the target to be tracked according to the changed attributes, and send the updated picture of the target to be tracked to the monitoring front-end device associated with the route of the target to be tracked.

[0086] For example, when the monitoring front-end device determines that the blurred image successfully matches the moving target in the above manner, the image of the moving target may be reported to the central server.

[0087] When the central server receives the image of the mobile target reported by the monitoring front-end device, it can determine whether the attribute of the target to be tracked has changed based on the received image and the pre-saved image of the target to be tracked.

[0088] For example, the color attribute of the target to be tracked may change due to wearing or taking off clothes.

[0089] When the central server determines that the attributes of the target to be tracked have changed based on the picture of the mobile target, the picture of the target to be tracked can be updated based on the changed attributes. For example, the picture of the mobile target with changed attributes is used as the updated picture of the target to be tracked, and the updated picture of the target to be tracked is sent to the monitoring front-end device associated with the route of the target to be tracked, so that the monitoring front-end device can identify and track the target to be tracked based on the updated picture of the target to be tracked.

[0090] In order to enable those skilled in the art to better understand the technical solutions provided by the embodiments of the present application, the technical solutions provided by the embodiments of the present application are described below with reference to specific examples.

[0091] In this embodiment, the monitoring front-end device is taken as an example as a pan-tilt camera (referred to as camera).

[0092] In this embodiment, the camera uses the latitude and longitude function or radar technology to measure the distance between the moving target and the camera, and determines whether the moving target is within the effective recognition range. Different recognition strategies are used to identify and track the target based on the judgment result.

[0093] For example, for a moving target outside the effective recognition range, different methods can be used to blur the details of the image of the target to be tracked according to the distance between the moving target and the camera.

[0094] For example, at different distances, the target photo can be blurred to different levels of detail by reducing the pixels of the target photo, using a blur algorithm (such as a Gaussian algorithm), etc., and the blurred photo and the current moving target are matched for similarity. After a successful match, the target information and the photo of the moving target are reported to the central server.

[0095] The central server collects, analyzes and judges the information reported by different cameras in an automatic or manual intervention mode, and forms new tracking instructions based on the judged information to guide the camera's further tracking operations.

[0096] For example, the moving target may include a vehicle, a person, or an animal, and the camera may detect the moving target in a manner that includes vehicle recognition, intrusion movement recognition, and the like.

[0097] The video tracking solution in this embodiment is described below.

[0098] 1. Installation and configuration: Install the PTZ cameras and connect them to each other or to the central server.

[0099] For example, if the central server is a specially deployed server, each pan-tilt camera establishes a network connection with the central server respectively, and saves the camera's static parameters such as GPS information, altitude, horizontal distance between cameras, etc. to the central server; if the central server is a pan-tilt camera, each pan-tilt camera establishes a network connection with each other, saves the camera's static parameters, and elects a central server.

[0100] 2. Initiate a video patrol task: Configuration parameters for a video patrol task can include a photo of the target to be tracked, patrol time (if not configured, it defaults to all-day), patrol range (such as patrol direction, which defaults to the entire range if not configured), patrol mode (long-range or close-range), and salient features to be fuzzy matched (such as outlines, colors, etc., i.e., the designated salient features mentioned above). Once the video patrol task configuration is complete, the video patrol task can be issued to each camera.

[0101] 3. Matching when a moving target appears: When a camera detects a moving target within its patrol range, it can use radar or multiple cameras linked by GPS to measure the distance between the target and the camera and compare it with the camera's effective recognition range. If the distance between the target and the camera exceeds the camera's effective recognition range, fuzzy matching can be used for target recognition. For example, previously issued photos of the target to be tracked can be blurred by reducing the pixel count or using a fuzzy algorithm (such as a Gaussian algorithm) to reduce the target's detailed information and extract salient features such as outline and color.

[0102] For example, detail blur and distance are calculated using a linear algorithm:

[0103] y=ax+b

[0104] Among them, x is the distance between the moving target and the camera, y is the blur level, such as the blur radius of the Gaussian algorithm, and a and b are empirical constants.

[0105] For example, the salient features of the target to be tracked may be matched with the salient features of the current moving target to calculate the comprehensive similarity.

[0106] For example, a weighted average algorithm can be used for multiple significant features:

[0107]

[0108] Among them, f i is the similarity of each significant feature, n is the number of significant features, f is the weighted similarity (i.e., comprehensive similarity), a i is the weighting coefficient.

[0109] If the comprehensive similarity between the salient features of the current moving target and the salient features of the target to be tracked exceeds the set similarity threshold (such as the first similarity threshold mentioned above), the match is determined to be successful, and the target information of the current moving target and the photo of the current moving target (such as a photo captured in real time) are reported to the central server, and the current moving target is tracked.

[0110] For example, in a multi-target tracking scenario, the target with the highest comprehensive similarity within the patrol range may be selected for tracking.

[0111] 4. Capture moving targets and perform precise matching: When a moving target moves into the effective recognition range of the camera, structured detail matching can be used. For example, details such as human movements, animal forms, license plates and headlights can be matched. If the similarity exceeds the set similarity threshold (such as the second similarity threshold mentioned above), the match is determined to be successful, the target information and the photo of the moving target are reported to the central server, and the current moving target is tracked.

[0112] 5. The central server makes a decision based on target information: The central server can summarize and analyze the target information reported by each camera, such as the similarity of each significant feature, the similarity of each detailed feature, location information, altitude confidence, movement speed, and other information. The central server can also describe the routes that the moving target has already traveled and the possible routes (i.e., candidate routes). Based on the candidate route with the highest credibility, the central server predicts the most likely route of the moving target and pre-dispatches each camera on the route into patrol mode to wait for the target to be tracked to enter the field of view.

[0113] For example, the candidate travel route with the highest credibility may be highlighted and checked by relevant personnel.

[0114] For example, if the target to be tracked has a property change, such as a color change, the photo of the target to be tracked can be changed, and the updated photo of the target to be tracked can be used.

[0115] 6. End inspection: The inspection ends when the end task command is received or the task inspection end time is reached.

[0116] It can be seen that the video tracking solution of this embodiment has at least the following beneficial effects:

[0117] 1) The capture capability is improved through the cooperation of fuzzy feature matching and detail feature matching.

[0118] 2) Predict and lock the target position in advance through fuzzy matching, making it easier to link other cameras for preparation and adjustment of shooting.

[0119] 3) Solve the problem of tracking the target when it exceeds the effective visual range.

[0120] 4) The results of fuzzy matching serve as a preliminary perception guide for precise matching.

[0121] 5) Through comprehensive analysis and judgment of information from multiple PTZ cameras, the system can automatically and continuously capture, track, and lock onto moving targets in real time.

[0122] The above describes the method provided by this application. The following describes the device provided by this application:

[0123] See Figure 2 , is a structural diagram of a video tracking device provided in an embodiment of the present application, such as Figure 2 As shown, the video tracking device may include:

[0124] The receiving unit 210 is configured to receive a video inspection task, wherein the video inspection task includes image information of a target to be tracked;

[0125] The determining unit 220 is configured to determine the distance between the moving target and the monitoring front-end device when a moving target is detected to be within the inspection range;

[0126] A matching unit 230 is configured to perform detail blurring on the image of the target to be tracked, and match the target based on the blurred image, when the distance between the target and the monitoring front-end device exceeds the effective recognition range of the monitoring front-end device;

[0127] The tracking unit 240 is configured to track the moving target if the matching unit successfully completes the matching.

[0128] In some embodiments, the matching unit 230 performs detail blurring processing on the target image to be tracked, including:

[0129] Reducing the pixels of the target image to be tracked;

[0130] And / or, using a fuzzy algorithm on the target image to be tracked.

[0131] In some embodiments, the matching unit 230 performs detail blurring processing on the target image to be tracked, including:

[0132] Determining, based on the distance between the moving target and the device, a blur level that matches the distance;

[0133] According to the blur level, detail blurring processing is performed on the target image to be tracked.

[0134] In some embodiments, the matching unit 230 matches the moving target based on the blurred image, including:

[0135] Extracting significant features from the blurred image to obtain significant features of the target to be tracked; and extracting significant features from the moving target to obtain significant features of the moving target;

[0136] Similarity matching is performed between the salient features of the target to be tracked and the salient features of the moving target.

[0137] In some embodiments, the video inspection task further includes designated significant features to be fuzzy matched;

[0138] The matching unit 230 extracts significant features from the blurred image to obtain significant features of the target to be tracked; and extracts significant features from the moving target to obtain significant features of the moving target, including:

[0139] Performing designated salient features extraction on the blurred image to obtain designated salient features of the target to be tracked; and performing designated salient features extraction on the moving target to obtain designated salient features of the moving target.

[0140] In some embodiments, the video inspection task further includes an inspection mode, and the inspection mode includes a long-range inspection or a close-range inspection;

[0141] Different inspection modes correspond to different effective recognition ranges.

[0142] See Figure 3 , is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. The electronic device may include a processor 301 and a memory 302 storing machine-executable instructions. The processor 301 and the memory 302 may communicate via a system bus 303. Furthermore, by reading and executing the machine-executable instructions corresponding to the video tracking control logic in the memory 302, the processor 301 may perform the video tracking method described above.

[0143] The memory 302 mentioned herein may be any electronic, magnetic, optical, or other physical storage device that may contain or store information, such as executable instructions, data, and the like. For example, the machine-readable storage medium may be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, a storage drive (such as a hard disk drive), a solid-state drive, any type of storage disk (such as a CD, DVD, etc.), or similar storage media, or a combination thereof.

[0144] In some embodiments, a machine-readable storage medium is also provided. Figure 3 The memory 302 in the machine-readable storage medium stores machine-executable instructions. When the machine-executable instructions are executed by the processor, the video tracking method described above is implemented. For example, the machine-readable storage medium can be a ROM, RAM, CD-ROM, magnetic tape, floppy disk, or optical data storage device.

[0145] See Figure 4, is a structural diagram of a video tracking system provided in an embodiment of the present application, such as Figure 4 As shown, the video tracking system may include: a monitoring front-end device 410 and a central server 420; wherein:

[0146] The central server 420 is used to send a video inspection task to the monitoring front-end device 410, wherein the video inspection task includes image information of the target to be tracked;

[0147] The monitoring front-end device 410 is used to determine the distance between the moving target and the device when a moving target is detected to appear within the patrol range; when the distance between the moving target and the device exceeds the effective recognition range of the device, blur the details of the image of the target to be tracked, match the moving target based on the blurred image, and track the moving target if the match is successful.

[0148] In some embodiments, the monitoring front-end device 410 is further configured to report the target information of the mobile target to the central server 420 if the match is successful;

[0149] The central server 420 is further configured to predict the travel route of the target to be tracked based on the received target information, and to perform patrol scheduling on the monitoring front-end devices associated with the travel route of the target to be tracked based on the travel route of the target to be tracked.

[0150] In some embodiments, the central server 420 is specifically configured to predict candidate routes of the target to be tracked based on the received target information, and determine the candidate route with the highest credibility as the route of the target to be tracked.

[0151] In some embodiments, the monitoring front-end device 410 is further configured to report the image of the moving target to the central server if the match is successful;

[0152] The central server 420 is also used to update the image of the target to be tracked according to the changed attributes when it is determined based on the image of the mobile target that the attributes of the target to be tracked have changed, and send the updated image of the target to be tracked to the monitoring front-end device associated with the route of the target to be tracked.

[0153] Exemplarily, the central server 420 may include a server other than the multiple monitoring front-end devices 410 for performing video tracking tasks (such as a background server, also referred to as a central platform) or one of the multiple monitoring front-end devices 410 for performing video tracking tasks (also referred to as a judgment server).

[0154] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0155] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A video tracking method, applied to monitoring front-end equipment, characterized in that: The method comprises: Receive a video inspection task, wherein the video inspection task includes image information of a target to be tracked; wherein the image information of the target to be tracked included in the video inspection task includes the image of the target to be tracked or the storage address of the image of the target to be tracked; When a moving target is detected within the inspection range, determining the distance between the moving target and the device; When the distance between the mobile target and the device exceeds the effective recognition range of the device, the details of the image of the target to be tracked are blurred, the mobile target is matched based on the blurred image, and if the match is successful, the mobile target is tracked.

2. The method according to claim 1, characterized in that The performing detail blurring processing on the target image to be tracked includes: Reducing the pixels of the image of the target to be tracked; And / or, a fuzzy algorithm is used to perform detail fuzzy processing on the image of the target to be tracked.

3. The method according to claim 1, characterized in that The performing detail blurring processing on the target image to be tracked includes: Determining, based on the distance between the moving target and the device, a blur level that matches the distance; According to the blur level, detail blurring processing is performed on the target image to be tracked.

4. The method according to claim 1, wherein The matching of the moving target according to the blurred image includes: Extracting significant features from the blurred image to obtain significant features of the target to be tracked; and extracting significant features from the moving target to obtain significant features of the moving target; Similarity matching is performed between the salient features of the target to be tracked and the salient features of the moving target.

5. The method according to claim 4, characterized in that The video inspection task also includes designated significant features that need to be fuzzy matched; Extracting significant features from the blurred image to obtain significant features of the target to be tracked; And, extracting significant features of the moving target to obtain significant features of the moving target, including: Extracting designated salient features from the blurred image to obtain designated salient features of the target to be tracked; Furthermore, designated significant features of the moving target are extracted to obtain the designated significant features of the moving target.

6. The method according to claim 1, characterized in that The video inspection task also includes an inspection mode, and the inspection mode includes a long-range inspection or a close-range inspection; Different inspection modes correspond to different effective recognition ranges.

7. A video tracking device, deployed in a monitoring front-end device, characterized in that: The device comprises: A receiving unit, configured to receive a video patrol task, wherein the video patrol task includes image information of a target to be tracked; wherein the image information of the target to be tracked included in the video patrol task includes the image of the target to be tracked or the storage address of the image of the target to be tracked; A determination unit, configured to determine the distance between the moving target and the monitoring front-end device when a moving target is detected to be present within the inspection range; a matching unit configured to, when the distance between the mobile target and the monitoring front-end device exceeds the effective recognition range of the monitoring front-end device, perform detail blurring processing on the image of the target to be tracked, and match the mobile target based on the blurred image; The tracking unit is used to track the moving target when the matching unit successfully matches.

8. The device according to claim 7, characterized in that The matching unit performs detail blurring processing on the target image to be tracked, including: Reducing the pixels of the target image to be tracked; and / or, using a fuzzy algorithm on the image of the target to be tracked; and / or, The matching unit performs detail blurring processing on the target image to be tracked, including: Determining, based on the distance between the moving target and the device, a blur level that matches the distance; Performing detail blurring processing on the image of the target to be tracked according to the blur level; and / or, The matching unit matches the moving target according to the blurred image, including: Extracting significant features from the blurred image to obtain significant features of the target to be tracked; and extracting significant features from the moving target to obtain significant features of the moving target; Performing similarity matching between the salient features of the target to be tracked and the salient features of the moving target; and / or, The video inspection task also includes designated significant features that need to be fuzzy matched; The matching unit extracts significant features from the blurred image to obtain significant features of the target to be tracked; and extracts significant features from the moving target to obtain significant features of the moving target, including: Extracting designated significant features from the blurred image to obtain designated significant features of the target to be tracked; and extracting designated significant features from the moving target to obtain designated significant features of the moving target; and / or, The video inspection task also includes an inspection mode, and the inspection mode includes a long-range inspection or a close-range inspection; Different inspection modes correspond to different effective recognition ranges.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores machine-executable instructions that can be executed by the processor, and the processor is configured to execute the machine-executable instructions to implement the method according to any one of claims 1 to 6.

10. A machine-readable storage medium, characterized in that The machine-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

11. A video tracking system, characterized in that: include: Monitor front-end devices and central servers; including: The central server is used to issue a video patrol task to the monitoring front-end device, wherein the video patrol task includes image information of the target to be tracked; wherein the image information of the target to be tracked included in the video patrol task includes the image of the target to be tracked or the storage address of the image of the target to be tracked; The monitoring front-end device is used to determine the distance between the moving target and the device when a moving target is detected to appear within the patrol range; when the distance between the moving target and the device exceeds the effective recognition range of the device, blur the details of the image of the target to be tracked, match the moving target based on the blurred image, and track the moving target if the match is successful.

12. The system according to claim 11, wherein: The monitoring front-end device is further configured to report the target information of the mobile target to the central server if the match is successful; The central server is further configured to predict the travel route of the target to be tracked based on the received target information, and to perform patrol scheduling on the monitoring front-end devices associated with the travel route of the target to be tracked based on the travel route of the target to be tracked; The central server is specifically configured to predict candidate routes of the target to be tracked based on the received target information, and determine the candidate route with the highest credibility as the route of the target to be tracked; Wherein, the monitoring front-end device is further used to report the picture of the moving target to the central server when the matching is successful; The central server is further configured to update the image of the target to be tracked according to the changed attributes when it is determined based on the image of the mobile target that an attribute change has occurred to the target to be tracked, and to send the updated image of the target to be tracked to a monitoring front-end device associated with the route of the target to be tracked.

Images