A method and device for personnel tracking, a method and device for personnel filing

Abstract

The present invention provides a personnel tracking method and device, as well as a personnel filing method and device, belonging to the field of monitoring technology. The method includes: receiving a tracking loss message sent by a first monitoring device, where the tracking loss message at least includes the feature information of the tracked person; determining the next-level alternative monitoring device according to the first monitoring device, and determining the threshold adjustment information of the next-level alternative monitoring device; sending a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction at least includes the feature information of the tracked person and the threshold adjustment information; determining a second similarity threshold according to the threshold adjustment information. By using the method provided by the present invention, the accuracy of tracking personnel can be improved, and further the accuracy of personnel filing can be improved.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology, and in particular, to a personnel tracking method and device, and a personnel filing method and device. Background Art

[0002] With the rapid development of information technology, monitoring technology has been widely applied in various industries and places, providing many conveniences for people's daily work and life. However, there are also many problems in the single monitoring application mode. If you need to find the monitoring target, you still need to rely on manual real-time viewing and manual retrieval.

[0003] Since the monitoring devices in the monitoring scenario are often independent of each other, and the field of view of the monitoring devices is limited, it is impossible to perform continuous relay tracking on the monitoring target, and it is difficult to meet the needs of modern public security prevention and control applications.

[0004] How to correctly file the same monitoring target is also a technical problem that needs to be solved. The filing process of the existing technology generally performs face feature matching on the faces in each frame of the video to find the faces belonging to the same monitoring target. However, if only captured by the monitoring device, the angle, light, etc. of the face have a large impact on the face features. Simple feature matching and filing are prone to misfiling problems of classifying different faces into one file. Summary of the Invention

[0005] The present invention provides a personnel tracking method, a filing method and a device, which can realize intelligent automatic continuous tracking of the same monitoring target, and different monitoring devices have different similarity thresholds to improve the accuracy of tracking personnel and further improve the accuracy of filing.

[0006] The specific technical solutions provided by the embodiments of the present invention are as follows:

[0007] In a first aspect, an embodiment of the present invention provides a personnel tracking method, including:

[0008] Receiving a tracking loss message sent by a first monitoring device, where the tracking loss message at least includes the feature information of the tracked person;

[0009] Determining a next-level alternative monitoring device according to the first monitoring device, and determining the threshold adjustment information of the next-level alternative monitoring device;

[0010] Sending a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction at least includes the feature information of the tracked person and the threshold adjustment information;

[0011] Determining a second similarity threshold according to the threshold adjustment information;

[0012] Wherein, the next-level alternative monitoring device is configured to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction.

[0013] Optionally, in some embodiments, the threshold adjustment information is threshold adjustment information determined according to any one of the motion state of the tracked person obtained by the first monitoring device, the geographical distance between the second monitoring device in the next-level alternative monitoring device and the first monitoring device, and the estimated arrival time of the tracked person at the second monitoring device in the next-level alternative monitoring device.

[0014] Optionally, in some embodiments, determining the second similarity threshold according to the threshold adjustment information specifically includes:

[0015] Determining the second similarity threshold according to different arrival times of the tracked person and the threshold adjustment information; wherein, the different arrival times are different arrival times of the tracked person at the second monitoring device in the next-level alternative monitoring device.

[0016] Optionally, in some embodiments, determining the next-level alternative monitoring device according to the first monitoring device specifically includes:

[0017] Determining the next-level alternative monitoring device according to the geographical location of the first monitoring device.

[0018] Optionally, in some embodiments, determining the next-level alternative monitoring device according to the geographical location of the first monitoring device specifically includes:

[0019] Determining the monitoring devices within a preset distance range according to the geographical location of the first monitoring device as the next-level alternative monitoring devices.

[0020] Optionally, in some embodiments, determining the monitoring devices within a preset distance range according to the geographical location of the first monitoring device as the next-level alternative monitoring devices specifically includes:

[0021] The tracking loss message includes the motion state information of the tracked person;

[0022] Selecting the monitoring devices within the preset distance range corresponding to the motion state according to the motion state information as the next-level alternative monitoring devices.

[0023] Optionally, in some embodiments, the tracking loss message sent by the first monitoring device is specifically:

[0024] The tracking loss message sent by the first monitoring device when the tracked person does not meet the monitoring conditions of the first monitoring device.

[0025] Optionally, in some embodiments, the tracking loss message further includes the motion state information of the person being tracked;

[0026] The relay tracking instruction further includes: the estimated arrival time information for reaching the next-level alternative monitoring device;

[0027] Wherein, the estimated arrival time information is determined according to the geographical distance between the second monitoring device and the first monitoring device in the next-level alternative monitoring device, and the motion state information of the monitoring target.

[0028] Optionally, in some embodiments, it further includes:

[0029] Determine a second similarity threshold according to the estimated arrival time information and the threshold adjustment information.

[0030] Optionally, in some embodiments, the next-level alternative monitoring device is further configured to send a tracking success message for the person being tracked;

[0031] Wherein, the tracking success message is sent after the monitoring device determines that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and determines that the monitoring target is the person being tracked.

[0032] Optionally, in some embodiments, it further includes:

[0033] After receiving the tracking success message for the person being tracked, send a monitoring stop message for the person being tracked to each of the next-level alternative monitoring devices in the most recently determined next-level alternative monitoring devices except the monitoring device that currently sends the tracking success message.

[0034] Optionally, in some embodiments, the receiving the tracking loss message sent by the first monitoring device specifically includes:

[0035] According to the tracking end condition, confirm that when the tracking end condition is not met, receive the tracking loss message sent by the first monitoring device;

[0036] When the tracking end condition is met, send a tracking end instruction to the most recently determined next-level alternative monitoring device and do not receive the tracking loss message sent by the monitoring device.

[0037] In a second aspect, an embodiment of the present invention further provides a personnel filing method, which adopts the personnel tracking method described in the first aspect, and further includes:

[0038] Obtain the data information of the person being tracked. The data information is sent by the monitoring device after confirming that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and confirming that the monitoring target is the person being tracked, and at least includes the captured image of the person being tracked and the data information corresponding to the capture time.

[0039] According to the data information of the person being tracked, perform clustering analysis and file the captured images determined to be of the person being tracked.

[0040] Optionally, in some embodiments, the performing clustering analysis and filing the captured images determined to be of the person being tracked specifically includes:

[0041] At intervals of a set duration, perform clustering analysis on all the captured images determined to be of the person being tracked once, and file the captured images belonging to the person being tracked; or

[0042] After the interval duration without receiving the data information of the person being tracked exceeds the set threshold, perform clustering analysis on all the captured images determined to be of the person being tracked that have been received, and file the captured images determined to be of the person being tracked.

[0043] Optionally, in some embodiments, the filing the captured images determined to be of the person being tracked specifically includes:

[0044] Sort the corresponding monitoring devices that send the corresponding captured images in the order of the capture time corresponding to the filed captured images, and obtain the virtual movement track of the person being tracked composed of each monitoring device as a track node;

[0045] According to the actual geographical location information of each monitoring device in the virtual movement track, determine the unreasonable track nodes;

[0046] Delete the corresponding captured images reported by the unreasonable track nodes.

[0047] In a third aspect, an embodiment of the present invention further provides a personnel tracking device, including: a receiving module, configured to receive a tracking loss message sent by a first monitoring device, where the tracking loss message at least includes the feature information of the person being tracked;

[0048] A relay tracking module, configured to determine the next-level alternative monitoring device according to the first monitoring device and determine the threshold adjustment information of the next-level alternative monitoring device; determine the second similarity threshold according to the threshold adjustment information; where the next-level alternative monitoring device is configured to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction;

[0049] A sending module, configured to send a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction at least includes the feature information of the person to be tracked and the threshold adjustment information.

[0050] Fourthly, an embodiment of the present invention provides a personnel filing device, which adopts the personnel tracking device described in the third aspect, and further includes:

[0051] An obtaining module, configured to obtain the data information sent by the personnel tracking device, where the data information is sent by the monitoring device of the tracking device after confirming that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and confirming that the monitoring target is the person to be tracked, and at least includes the captured image of the person to be tracked and the data information corresponding to the capture time;

[0052] A filing module, configured to perform clustering analysis according to the data information of the person to be tracked and file the captured images determined to be of the person to be tracked.

[0053] In the embodiment of the present invention, after receiving the tracking loss message sent by the monitoring device, the next-level alternative monitoring device is determined, and the second similarity threshold is determined according to the threshold adjustment information. The next-level alternative monitoring device tracks the person to be tracked according to the second similarity threshold. The process of determining the second similarity threshold in this technical solution is a dynamic determination process. The determined next-level alternative monitoring device may have different second similarity thresholds at different times, and different monitoring devices may also have different second similarity thresholds at the same time. The second similarity threshold can be reasonably allocated according to the actual environment and the feature information of the person to be tracked, so that the monitoring device can better track the person to be tracked and prevent missed tracking. Description of the Drawings

[0054] Figure 1 It is a schematic diagram of a monitoring system architecture provided by an embodiment of the present invention;

[0055] Figure 2 It is a schematic diagram of monitoring devices distributed in a certain area provided by an embodiment of the present invention;

[0056] Figure 3 It is a flowchart of a personnel tracking method provided by an embodiment of the present invention;

[0057] Figure 4 It is a signaling flowchart of a personnel tracking method provided by an embodiment of the present invention;

[0058] Figure 5 It is provided by an embodiment of the present invention and Figure 4 corresponding to a flowchart of a personnel tracking method;

[0059] Figure 6Flowchart of a tracking method for a monitoring device end provided by an embodiment of the present invention;

[0060] Figure 7 Flowchart of a tracking method for a server end provided by an embodiment of the present invention;

[0061] Figure 8 Schematic diagram of a positional relationship between monitoring devices provided by an embodiment of the present invention;

[0062] Figure 9 Another flowchart of a tracking method for a monitoring device end provided by an embodiment of the present invention;

[0063] Figure 10 Another flowchart of a tracking method for a server end provided by an embodiment of the present invention;

[0064] Figure 11 Another schematic diagram of a positional relationship between monitoring devices provided by an embodiment of the present invention. Detailed implementation manners

[0065] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0066] Figure 1 An exemplary system architecture applicable in the embodiments of the present invention is shown. In this system architecture, monitoring devices 101 - 199 and a server 201 are included.

[0067] The monitoring devices 101 - 199 are connected to the server 201 through a wireless network. The number of monitors here is only an example, and in actual applications, the number of monitors will vary; the monitoring devices are electronic devices with the function of collecting images, such as cameras, video cameras, video recorders, etc., and they collect monitoring video streams in real time. The network - side device is described by taking the server 201 as an example in this embodiment. The server 201 is a server or a server cluster, a cloud computing center or a control platform composed of several servers, etc., which communicates with each monitoring device and conducts data interaction, and can also view the information of the video streams in each monitoring device.

[0068] In the embodiments of the present invention, the next-level alternative monitoring devices refer to a set of one or more monitoring devices. The process of determining the next-level alternative monitoring devices by the server 201 is a dynamic determination process, or the monitoring devices within a preset distance range according to the geographical location of the monitoring device that sends the tracking loss message, or the monitoring device range determined according to the movement state of the tracked person, or the monitoring device range that can be reached at the same time is determined as the next-level alternative monitoring device. In this application and the following embodiments, the monitoring devices within a preset distance range are taken as examples for illustration, but should not be used as a limitation for defining the next-level alternative monitoring devices.

[0069] See Figure 2 as shown Figure 2 is a schematic diagram of the monitoring devices distributed in a certain area. Each dot in the figure represents a certain monitoring device. Each monitoring device is an independent device that does not communicate with each other, but only communicates with the server 201, that is, data interaction occurs between the monitoring device and the server. The distribution in the figure is a schematic distribution.

[0070] Here, the dynamic process of how the server 201 determines the next-level alternative monitoring devices is described, and other processes are briefly described.

[0071] Suppose that when the server 201 sends the characteristic information of the tracked person to the monitoring devices in the current area, the monitoring devices in the current area search for monitoring targets with similarity meeting the preset conditions in the local monitoring video stream according to the characteristic information of the tracked person. Suppose the monitoring device 4 successfully monitors the tracked person. At this time, the monitoring device 4 is the first monitoring device, generates a corresponding tracking identifier for the current tracked person and reports it to the server. When the tracked person leaves the monitoring area, the monitoring device 4 sends a tracking loss message of the current tracked person to the server 201.

[0072] When the tracking end condition is not met, the server 201 determines the monitoring devices within a preset distance range as the next-level alternative monitoring devices according to the geographical location of the monitoring device 4 that sends the tracking loss message. The set A simply shows the next-level alternative monitoring devices defined by the server 201 when the tracked person leaves the monitoring device 4. The set A contains the monitoring device 4, and of course, it may not contain the monitoring device 4. Those skilled in the art can make adaptive adjustments according to the actual situation; when selecting the next-level alternative monitoring devices, it is also possible not to select, that is, to determine all the monitoring devices in the network as the next-level alternative monitoring devices.

[0073] The server 201 sends a relay tracking instruction to the next-level alternative monitoring devices, that is Figure 2For each monitoring device 5, 12, 13, 14 in set A, whether it includes monitoring device 4 can be defined according to the situation. For the convenience of explanation here, temporarily do not regard the monitoring device 4 that discovers the tracked person as the next-level alternative monitoring device of the current set, that is, assume that there is no situation where the tracked person turns back. After each monitoring device in set A receives the relay tracking instruction sent by the server 201, according to the characteristic information of the tracked person in the relay tracking instruction, it is determined whether there is a monitoring target that meets the conditions in the local monitoring video stream.

[0074] Suppose that at this time, monitoring device 12 discovers the tracked person. When the tracked person leaves the monitoring area, that is, after the monitoring target disappears from the video screen, or when the tracking and monitoring conditions are not met, monitoring device 12 sends a message of lost tracking to server 201. After server 201 receives the lost tracking message, again according to the geographical location of the monitoring device 12 that sends the lost tracking message, the monitoring devices within the preset distance range are determined as the next-level alternative devices. Due to the change of geographical location, at this time, the next-level alternative monitoring devices form a new set B, as Figure 2 The set B shown in contains monitoring devices 13, 14, 16, 11, 15, and server 201 sends the relay tracking instruction to each monitoring device in set B.

[0075] When defining the next-level monitoring device, server 201 dynamically defines the next-level alternative monitoring device according to the geographical location of the monitoring device that sends the lost tracking instruction and within the preset distance range, or can also define the next-level alternative monitoring device according to other conditions. During this process, it is possible that the next-level alternative device defined last time will still be defined as the next-level alternative monitoring device when defined again, as Figure 2 The monitoring devices 13, 14 in set A shown in are also located in set B. Of course, according to the geographical location, there may also be monitoring devices that have already discovered the tracked person that will be defined as the next-level alternative monitoring device, such as monitoring device 4, but it is not shown in this embodiment.

[0076] After the monitoring devices in set B receive the relay tracking instruction, suppose that monitoring device 16 discovers that the tracked person appears in the monitoring video. When the tracked person disappears from the video screen, server 201 again defines the next-level alternative monitoring device according to the geographical location of monitoring device 16 and the preset distance, that is, as Figure 2 The set C shown in, set C includes monitoring devices 17, 18, 15, 19, 20, and so on repeatedly until the end tracking condition is met.

[0077] The following is a supplementary explanation of the determined set of the next-level alternative monitoring devices.

[0078] After the monitoring device 4 tracks the person to be tracked, when the person to be tracked leaves the monitoring area of the monitoring device 4, the server 201 determines the next-level alternative monitoring devices, that is, set A. When the monitoring device 12 in set A monitors the person to be tracked and the person to be tracked leaves the monitoring area of the monitoring device 12, the server 201 determines the next-level alternative monitoring devices, that is, set B. When the monitoring device 16 in set B monitors the person to be tracked and the person to be tracked leaves the monitoring area of the monitoring device 16, the server 201 determines the next-level alternative monitoring devices, that is, set C. Combined with Figure 2 , it can be seen that the server 201 determines the next-level alternative monitoring devices as a dynamic definition process, and can re-define different next-level alternative monitoring devices according to different determination conditions at different monitoring device locations, and this process is a repeated process until the tracking end condition is met.

[0079] To describe the technical solution of this application in detail, the following will be described in detail through specific embodiments.

[0080] Based on Figure 1 the system architecture shown, an embodiment of the present invention provides a personnel tracking method by taking tracking a person to be tracked as an example, as shown in Figure 3 shown, including:

[0081] S301. Receive the tracking loss message sent by the first monitoring device, and the tracking loss message includes at least the feature information of the person to be tracked;

[0082] When the first monitoring device sends a tracking loss message, it means that the person to be tracked has left the monitoring area, or cannot match the similarity in the monitored picture and does not meet the monitoring conditions. The first monitoring device sends a tracking message to the server 201, and the tracking loss message includes the feature information of the person to be tracked.

[0083] S302. Determine the next-level alternative monitoring devices according to the first monitoring device, and determine the threshold adjustment information of the next-level alternative monitoring devices;

[0084] After receiving the loss message, determine the next-level alternative monitoring devices. The next-level alternative monitoring devices can be defined within a preset range, or the monitoring devices preset according to the movement state of the person to be tracked, or determined as the next-level alternative monitoring devices according to the monitoring device range that the person to be tracked can reach at the same time. There is no limitation here.

[0085] S303. Send a relay tracking instruction to the next-level alternative monitoring devices, and the relay tracking instruction includes at least the feature information of the person to be tracked and the threshold adjustment information;

[0086] After determining the threshold adjustment information for the next-level alternative monitoring devices, send a relay tracking instruction to the determined next-level alternative monitoring devices. The next-level alternative monitoring devices are a set of alternative monitoring devices, which may include one monitoring device or multiple monitoring devices. Correspondingly, the threshold adjustment information can be the same, different, or partially the same for different next-level alternative monitoring devices. The next-level alternative monitoring devices are used to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction. The second similarity threshold is related to the threshold adjustment information.

[0087] S304. Determine the second similarity threshold according to the threshold adjustment information;

[0088] The process of determining the second similarity threshold can be performed by the server side and then sent to the next-level alternative monitoring devices after determination, or can be executed by the received next-level alternative monitoring devices to determine the second similarity threshold according to the threshold adjustment information. The execution entity is not limited here. For those skilled in the art, whether the execution entity is the server 201 or the next-level alternative monitoring devices, its technical solution can be implemented.

[0089] In the embodiment of the present invention, after receiving the tracking loss message sent by the monitoring device, determine the next-level alternative monitoring devices and determine the second similarity threshold according to the threshold adjustment information. The next-level alternative monitoring devices track the person to be tracked according to the second similarity threshold. The process of determining the second similarity threshold in this technical solution is a dynamic determination process. The determined next-level alternative monitoring devices may have different second similarity thresholds at different times, different monitoring devices may have different second similarity thresholds at the same time, and the same monitoring device may also have different second similarity thresholds for different persons to be tracked at the same time. The second similarity threshold can be reasonably allocated according to the actual environment and the characteristic information of the person to be tracked, so that the monitoring device can better track the person to be tracked and prevent missed tracking.

[0090] Optionally, as one of the embodiments, the threshold adjustment information is determined according to at least one of the motion state of the person to be tracked obtained by the first monitoring device, the geographical distance between the second monitoring device in the next-level alternative monitoring devices and the first monitoring device, and the preset arrival time of the person to be tracked at the second monitoring device in the next-level alternative monitoring devices.

[0091] Determining the second similarity threshold according to the threshold adjustment information specifically includes:

[0092] Determine the second similarity threshold according to the different arrival times of the person to be tracked and the threshold adjustment information; wherein, the different arrival times are the different arrival times of the person to be tracked at the second monitoring device in the next-level alternative monitoring devices.

[0093] The following is an illustration through specific embodiments.

[0094] Refer to Table 1, Table 2, and Table 3.

[0095] Table 1 is a comparison table of thresholds and time information, Table 2 is a comparison table of thresholds and distance information, and Table 3 is a comparison table of thresholds and motion state information.

[0096] Table 1 - Comparison Table of Thresholds and Time Information

[0097]

[0098] Table 2 - Comparison Table of Thresholds and Distance Information

[0099]

[0100]

[0101] Table 3 - Comparison Table of Thresholds and Motion State Information

[0102] Motion state Threshold Walking state 0.9 Running state 0.88 Cycling state 0.86 Driving a motor vehicle 0.85

[0103] The above information is exemplary information. Those skilled in the art should understand that the thresholds and the corresponding preset arrival time information, distance information, and motion state information can be adjusted accordingly according to the actual situation.

[0104] Table 1 is a comparison table of thresholds and time information.

[0105] Estimating the arrival time of the tracked person at the second monitoring device among the next-level alternative monitoring devices means the time when the tracked person leaves the first monitoring device and arrives at the second monitoring device among the next-level alternative monitoring devices. Assume X is the time when the tracked person leaves the first monitoring device. Within 3 minutes after time X, the next-level alternative monitoring device uses a threshold of 0.85 as the second similarity threshold. Within 3 - 5 minutes after time X, the next-level alternative monitoring device uses a threshold of 0.86 as the second similarity threshold. Within 5 - 10 minutes after time X, the next-level alternative monitoring device uses a threshold of 0.87 as the second similarity threshold. By analogy, when within 20 - 30 minutes after time X, the second similarity threshold of 0.90 is used. For similarity thresholds outside other time ranges, the first similarity threshold can be used to define.

[0106] Table 2 is a comparison table of thresholds and distance information.

[0107] As shown in Table 2, the second similarity threshold information is determined according to the geographical distance between the second monitoring device and the first monitoring device in the next-level alternative monitoring devices. The server 201 knows the specific location information of each monitoring device. When the tracked person leaves the first monitoring device, different thresholds are determined according to the distance between the second monitoring device and the first monitoring device in the next-level alternative monitoring devices. The aforementioned next-level alternative monitoring devices are a set of monitoring devices and may include multiple monitoring devices. Therefore, the geographical distance between each monitoring device in the next-level alternative monitoring devices and the first monitoring device is different. When the geographical distance between the second monitoring device and the first monitoring device in the next-level alternative monitoring devices does not exceed 100 m, the second similarity threshold of 0.85 is adopted. When the geographical distance between the third monitoring device and the first monitoring device in the next-level alternative monitoring devices is between 100 m and 500 m, the second similarity threshold of 0.86 is adopted. By analogy, when the geographical distance between the Nth monitoring device and the first monitoring device in the next-level alternative monitoring devices is between 3000 m and 5000 m, the second similarity threshold of 0.90 is adopted. For the similarity threshold outside the 5000 m distance range, the first similarity threshold can be used to define it.

[0108] Table 3 is a comparison table of thresholds and motion state information

[0109] As shown in Table 3, the server 201 can obtain the motion state information of the tracked person in the first monitoring device and can determine different thresholds according to the motion state information of the tracked person. If the tracked person is in a walking state, the second similarity threshold of 0.9 is adopted for the next-level alternative monitoring devices. If the tracked person is in a running state, the second similarity threshold of 0.88 is adopted for the next-level alternative monitoring devices. By analogy, if the tracked person is in a state of driving a motor vehicle, the second similarity threshold of 0.85 is adopted for the next-level alternative monitoring devices. For the similarity threshold of other motion states, the first similarity threshold can be used to define it.

[0110] Of course, the second similarity threshold can also be determined by any combination of the above two methods or the combination of the three methods. When both situations or all three situations are satisfied, the determination rule of the threshold can be defined according to the actual situation, and any one of the highest threshold, the lowest threshold, and the median threshold can be adopted, or the average value method of the threshold can be adopted, or other rule definition methods. This is not limited here. It can be understood that one form of the threshold adjustment information is in the form of a table, and it can also be in a non-table form, and the specific form is not limited.

[0111] The following is an example of selecting the second similarity threshold from the thresholds in an alternative way. Those skilled in the art can define the threshold determination rule according to the actual situation.

[0112] For example, when the tracked person leaves the first monitoring device and is in a walking state, according to Table 3, the second similarity threshold of the next-level alternative monitoring device is 0.9. If the geographical distance between the second monitoring device and the first monitoring device among the next-level alternative monitoring devices is 1000m, then according to Table 2, the second similarity threshold is 0.87. At this time, it is assumed that within 3 minutes after the tracked person leaves the first monitoring device, according to Table 1, the second similarity threshold is 0.85. Then, according to the three different second similarity thresholds of 0.85, 0.87, and 0.9 in Table 1, Table 2, and Table 3, the second similarity threshold of the second monitoring device can be determined with the highest threshold, such as 0.9, or with the lowest threshold, such as 0.85, or with the middle threshold, such as 0.87.

[0113] If the second monitoring device does not detect the tracked person appearing in the local video stream within 3 minutes after the tracked person leaves the first monitoring device, then according to the threshold and time information comparison table in Table 1, the threshold will be adjusted to the threshold information for 3 minutes - 5 minutes, that is, 0.86. At this time, the thresholds of the second monitoring device according to Table 1, Table 2, and Table 3 are changed to 0.86, 0.87, and 0.90. If the lowest threshold is used as the second similarity threshold, then within 3 minutes - 5 minutes after the tracked person leaves the first monitoring device, the second similarity threshold is 0.86. If the highest threshold or the middle threshold is used as the second similarity threshold, the second similarity threshold remains unchanged.

[0114] And so on. If the second monitoring device still does not detect the tracked person appearing in the local video stream within the time of 3 minutes - 5 minutes, then the threshold that conforms to Table 1 will be changed to 0.87 within 5 minutes - 10 minutes, and the thresholds that conform to Table 2 and Table 3 remain unchanged, still being 0.87 and 0.90. According to the corresponding second threshold selection rule, one of 0.87, 0.87, and 0.90 is selected as the second similarity threshold.

[0115] It should be noted that the threshold determined in Table 1 will be adjusted over time. Therefore, the second similarity threshold is a dynamic process that changes with time. When determining the second monitoring device of the next-level alternative monitoring device after the tracked person leaves the first monitoring device in Table 2, the distance between the second monitoring device and the first monitoring device is basically fixed and does not change. Therefore, the second monitoring device and the first monitoring device are also relatively fixed thresholds. In Table 3, when the motion state of the tracked person is determined, the threshold is also relatively fixed. Therefore, in actual use, those skilled in the art can set reasonable threshold determination rules according to actual needs.

[0116] In the embodiments of the present application, different threshold adjustment information is determined through different parameters, so as to determine different second similarity thresholds for different monitoring devices in the next-level alternative monitoring devices. The process of determining the second similarity threshold is a dynamic determination process, comprehensively considering various factors that may affect the similarity threshold, and using the factors affecting similarity as a parameter for setting the second similarity threshold. The second similarity threshold of the monitoring device can be set more reasonably according to different environments and different factors, improving the success rate of tracking. Compared with a single similarity threshold, the setting is more flexible and can be suitable for tracking in various situations.

[0117] Optionally, as one of the embodiments, determining the next-level alternative monitoring devices according to the first monitoring device specifically includes: determining the next-level alternative monitoring devices according to the geographical location of the first monitoring device.

[0118] Determining the next-level alternative monitoring devices according to the geographical location of the first monitoring device specifically includes:

[0119] According to the geographical location of the first monitoring device, the monitoring devices within a preset distance range are determined as the next-level alternative monitoring devices.

[0120] The preset distance can be directly set as a range. For example, taking the monitoring device that currently sends the loss message as the center and forming a range with a preset M meters as the radius, where M is a positive integer. The monitoring devices within this range are the next-level alternative monitoring devices. The number of alternative monitoring devices is from one to several, which is not limited here. For example, monitoring device 102, monitoring device 103, monitoring device 104, etc., that is, in the form of the aforementioned set; it can also be the alternative target monitoring devices that can be reached within the movement range in the travel route where the monitoring target is located, and screening the X target monitoring devices with the shortest distance to the monitoring device that sends the loss message on all travel routes to each monitoring device, where X is a positive integer. For example, monitoring device 102, monitoring device 105, etc. No matter which method, the monitoring devices within the preset distance range can be determined as the next-level alternative monitoring devices through the preset distance range. This process is a dynamic determination process, and the next-level alternative monitoring devices can be determined according to the geographical location of the monitoring device that sends the loss message and the preset distance range.

[0121] Optionally, as one of the embodiments, the tracking loss message sent by the first monitoring device is specifically:

[0122] The tracking loss message sent by the first monitoring device when the person being tracked no longer meets the monitoring conditions of the first monitoring device.

[0123] When the monitoring conditions of the first monitoring device are not met, for example, when the tracked person leaves the monitoring area, or when the tracked person does not leave the monitoring area, but the pixels of the monitoring screen are not clear, resulting in the inability to identify the tracked person, or when there is a sudden interference resulting in no picture, etc. When the monitoring conditions of the first monitoring device are not met, the first monitoring device sends a tracking loss message to the server 201 so that the server can take the next action.

[0124] Optionally, as one of the embodiments, the tracking loss message further includes the motion state information of the tracked person;

[0125] The relay tracking instruction further includes: the estimated arrival time information to reach the next-level alternative monitoring device;

[0126] Among them, the estimated arrival time information is determined according to the geographical distance between the second monitoring device and the first monitoring device in the next-level alternative monitoring device, and the motion state information of the monitoring target.

[0127] Determine the second similarity threshold according to the estimated arrival time information and the threshold adjustment information.

[0128] The following is illustrated by specific embodiments.

[0129] Figure 8 It is a schematic diagram of a specific positional relationship between monitoring devices;

[0130] Among them, the dotted connecting line is the route that the tracked person can travel. It is assumed that the monitoring devices 121, 122, and 123 are independent of each other, do not perform data interaction, and only perform data interaction with the server 201.

[0131] The distance between the monitoring device 121 and the monitoring device 122 is 1000 meters, and the distance between the monitoring device 123 and the monitoring device 121 is 2000 meters.

[0132] Suppose the monitoring device 121 receives the feature information of the tracked person sent by the server 201, executes Figure 6 the process, and has tracked the tracked person. Currently, the tracked person is in a walking motion state with a speed of 2 m / s.

[0133] When the tracked person leaves the monitoring area of the monitoring device 121, the server 201 receives the tracking loss message sent by the monitoring device 121 and the information that the tracked person is in a walking motion state with a speed of 2 m / s. Under the condition that the tracking end condition is not met, the server 201 determines the monitoring devices within the preset distance range as the next-level alternative monitoring devices according to the geographical location of the monitoring device 121.

[0134] If it is assumed that the preset distance range is set to 2500m, then the server 201 will determine the monitoring devices 122 and 123 as the next-level alternative monitoring devices. According to the speed of the tracked person and the distance between the monitoring device 121 and the monitoring device 122, it is estimated that the time for the tracked person to travel from the monitoring device 121 to the monitoring device 122 is 500s. In practice, the estimated time is not equal to the actual arrival time, and there will be a certain error between the two. The error time range can be defined by itself. Here, the error time range is set as the adjustment time.

[0135] Here, it can be assumed that the set adjustment time is 10 minutes. Within the adjustment time range, the threshold is adjusted as shown in Table 4.

[0136] Table 4 - Corresponding Table of Estimated Arrival Time and Threshold

[0137] Estimated time of arrival X Threshold <X±1min 0.85 <X±2min 0.86 <X±3min 0.87 X ± 4 min 0.88 <X ± 5 min 0.89

[0138] Of course, the adjustment time can also be set as a dynamic value and adjusted appropriately according to the length of the estimated arrival time. If it arrives after 1 hour, the adjustment time is 10 minutes. If it arrives within 30 minutes, the adjustment time can be set to 5 minutes, or it can also be determined according to the estimated arrival time. For example: divide the estimated arrival time by N, where N is an integer greater than 1. Suppose the estimated arrival time is 1 hour and N is taken as 6, then the adjustment time is 10 minutes. If the estimated arrival time is 2 hours, the adjustment time is 20 minutes. Because the error of the estimated time will increase with the increase of time and distance, the adjustment time will increase accordingly.

[0139] The adjustment time is not limited here, and those skilled in the art can adjust it according to the actual situation.

[0140] Table 4 lists a situation of the threshold within the adjustment time range. When there is more than one condition satisfied for the estimated arrival time, the lowest threshold is executed.

[0141] Suppose the estimated time arrives after 15 minutes. Then, within the time range from 14 minutes to 16 minutes after the current time, when all conditions are satisfied, the lowest threshold of 0.85 is executed, that is, 0.85 is used as the second similarity threshold. Within the time range from 13 minutes to 14 minutes and 16 minutes to 17 minutes after the current time, the lowest threshold is 0.86, that is, 0.86 is used as the second similarity threshold, and so on. Within the time range from 10 minutes to 11 minutes and 19 minutes to 20 minutes after the current time, the threshold is 0.89, that is, 0.89 is used as the second similarity threshold. Outside other times, the unified first similarity threshold can be executed.

[0142] The estimated time of arrival can also be divided proportionally, and thresholds can be assigned to each divided interval in the form of proportional division. The proportional division can be preset to be N parts. For example, if the estimated time of arrival is 1 hour later, assuming N is 6, if it is equally divided, each divided area is 10 minutes. The threshold within the estimated time of arrival area can be the lowest, and the thresholds of other adjacent areas gradually increase. Assuming the estimated time is 1 hour, within the time range of 55 minutes - 65 minutes, the threshold is 0.85, and within the time ranges of 45 minutes - 55 minutes and 65 minutes - 75 minutes, the threshold is 0.86.

[0143] The above embodiments can also be referred to, and reasonable thresholds can be set in combination with the estimated time of arrival, and the second similarity threshold can be determined according to the thresholds. After those skilled in the art determine the estimated time of arrival information, they can determine the threshold adjustment information according to actual needs. The rule of the threshold adjustment information can be defined by themselves, and then the second similarity threshold can be determined in combination with the estimated time of arrival and the threshold adjustment information.

[0144] The technical solution of the embodiment of the present application determines the second similarity threshold through the determined estimated time of arrival information and threshold adjustment information. By predicting the arrival time of the tracked person and setting reasonable threshold adjustment information, the second similarity threshold is obtained, avoiding mis-tracking caused by reducing the threshold within an unreasonable time range, and relatively improving the success rate of tracking.

[0145] Preferably, in some of the following embodiments, the next-level alternative monitoring device is further configured to send a tracking success message of the tracked person;

[0146] Among them, the tracking success message is sent after the monitoring device determines that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and determines that the monitoring target is the tracked person.

[0147] After receiving the tracking success message of the tracked person, a monitoring stop message of the tracked person is sent to each of the next-level alternative monitoring devices in the next-level alternative monitoring devices determined most recently, except for the monitoring device that currently sends the tracking success message.

[0148] Receiving the tracking loss message sent by the first monitoring device specifically includes:

[0149] According to the tracking end condition, when it is confirmed that the tracking end condition is not met, receiving the tracking loss message sent by the first monitoring device; and / or,

[0150] When the tracking end condition is met, a tracking end instruction is sent to the next-level alternative monitoring device determined most recently, and the tracking loss message sent by the monitoring device is not received.

[0151] The following will be described through various embodiments.

[0152] As Figure 4 and Figure 5 shown Figure 4 is a signaling flowchart of a personnel tracking method Figure 5 and Figure 4 is the corresponding method flowchart. It should be noted that Figure 4 only an example of a process from the first monitoring device discovering the tracked person to the next-level alternative monitoring device tracking the tracked person is given, and it is not the entire process of this application. The entire process is a dynamic and repeated process until the tracking end condition is met. For details, please refer to Figure 5 .

[0153] The specific steps are as follows:

[0154] S401 / S5401: After the first monitoring device 101 determines that there is a monitoring target in the local monitoring video stream whose similarity meets the preset condition according to the feature information of the tracked person, it periodically obtains the captured images of the monitoring target and generates corresponding tracking identifiers;

[0155] When the monitoring device discovers that the monitoring target in the monitoring video stream meets the preset condition, that is, when the tracked person appears in the monitoring video, the first monitoring device that first tracks the monitoring target is defined as the first monitoring device 101 in this embodiment. The tracking identifier can be understood as a unique identifier used to represent each tracked person.

[0156] Since the first monitoring device 101 is the first monitoring device to discover the tracked person here, the preset condition can be the initial first similarity threshold or the defined second similarity threshold.

[0157] S402 / S5402: The first monitoring device 101 reports each captured image, the corresponding capture time, and the tracking identifier to the server 201;

[0158] The first monitoring device 101 reports the captured image to the server 201. When capturing the image, there is time information, which can be formed into a separate record file and reported to the server 201, or marked on the captured image, or reported to the server 201 in other ways, that is, reporting the data information to the server 201. The tracking identifier represents the information of the current tracked person. Therefore, the tracking identifier is reported together with the captured image. When the server 201 receives the image information and the tracking identifier reported by the monitoring device 101, it can determine that the captured picture information is the information of the tracked person.

[0159] S403 / S5403: After the first monitoring device 101 determines that the monitoring target has disappeared from the video screen, it sends a tracking loss message containing at least the tracking identifier to the server 201;

[0160] If the monitored target disappears from the video image, it indicates that the monitored target has left the monitored area or does not meet the tracking conditions. At this time, for the monitoring device, the person being tracked is in a state of lost tracking. Therefore, after the monitored target disappears from the video image, the monitoring device reports a message of lost tracking of the monitored target to the server 201.

[0161] S404 / S54041: The server 201 receives the lost tracking message.

[0162] S404 / S54042: The server 201 determines whether the tracking end condition is met.

[0163] If yes, execute step S5601; if no, execute step S54043.

[0164] The tracking end condition can be freely set or adjusted. For example, issuing an instruction to no longer track the current person being tracked, or ending the tracking when the person being tracked has not been tracked for a certain period of time, or exceeding a certain distance, etc. It can also be other end conditions, which are not limited here.

[0165] Of course, when the person being tracked has not been tracked, that is, when the person being tracked has not been tracked for a certain period of time, the reason needs to be investigated. It may be that the feature information of the person being tracked is incorrect or incomplete, resulting in the monitoring device being unable to match its features, or the estimated activity range of the person being tracked is inaccurate, resulting in the feature information of the person being tracked not being sent to the monitoring devices covered by the activity area of the person being tracked, etc.

[0166] S404 / S54043: According to the geographical location of the monitoring device that sends the lost tracking message, determine the monitoring devices within a preset distance range as the next-level alternative monitoring devices.

[0167] The network-side device is specifically described by the server 201 in this embodiment. After receiving the lost tracking message, the server 201 infers the approximate activity range of the person being tracked based on the geographical location of the monitoring device that sends the lost tracking message, and determines the monitoring devices within a preset distance range as the next-level alternative monitoring devices.

[0168] Regarding how to determine the next-level alternative monitoring devices, this embodiment takes the monitoring devices within a preset distance range as an example for illustration.

[0169] S405 / S5405: The server 201 sends a relay tracking instruction to the next-level alternative monitoring devices. The relay tracking instruction includes at least the feature information and tracking identifier of the person being tracked.

[0170] The tracking identifier is matched with the information of the person being tracked. When the next-level alternative monitoring device obtains the tracking identifier and the characteristic information of the person being tracked, it can match the obtained characteristic information of the person being tracked with the pedestrian information appearing in the monitoring video stream to monitor whether the person being tracked appears in the local monitoring video stream.

[0171] Moreover, the server 201 determines the tracking range by determining the next-level alternative devices. Not all monitoring devices participate in the relay tracking process. This method narrows the range of the next-level alternative monitoring devices, and the device receiving the tracking instruction can obtain the characteristic information of the person being tracked as soon as possible, reducing the possible missed tracking caused by the monitoring device receiving messages untimely.

[0172] S406 / S5406: After the next-level alternative monitoring device determines that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the characteristic information of the person being tracked, it periodically obtains the captured images of the monitoring target;

[0173] When the next-level alternative monitoring device finds that the monitoring target in the monitoring video stream meets the second similarity threshold, that is, it finds that the person being tracked appears in the monitoring video, it periodically obtains the captured images of the monitoring target. The second similarity threshold can be freely and flexibly set. For example, it can be the second similarity threshold, or the first similarity threshold that is uniformly preset when the condition of not meeting the second similarity threshold is adopted. Both are acceptable and are not limited here.

[0174] S407 / S5407: The next-level alternative monitoring device reports each captured image, the corresponding capture time, and the tracking identifier to the server 201;

[0175] The content executed in this step is the same as that in step S402 / S5402, and the execution entity is the monitoring device. See the description in step S402 / S5402.

[0176] S408 / S5408: After the next-level alternative monitoring device determines that the monitoring target has disappeared from the video screen, it sends a tracking loss message to the server 201;

[0177] Repeat the above steps S404 - 408, that is Figure 5 steps S54041 to S5408 in Figure 5 until the tracking end condition is met and the tracking stops, and execute

[0178] S409 / S5409: The server 201 performs clustering analysis based on the received captured images and archives the captured images determined to belong to the person being tracked.

[0179] After receiving the captured image, the server 201 performs clustering analysis. The clustering analysis can be performed periodically or according to the received clustering analysis instruction. This process is carried out in parallel with the tracking process. Even if the tracking process ends, clustering analysis can still be performed on the captured image, and the captured images belonging to the tracked person are archived.

[0180] In an embodiment of the present invention, the monitoring device performs target recognition based on the received feature information of the tracked person. If the tracked person is found, the monitoring target is periodically captured, and the captured image, the corresponding capture time, and the tracking identifier are reported to the server. When the tracked person leaves the monitoring area, after the monitoring device reports the message that the tracked person is lost to the server, the server determines the next-level alternative monitoring device for relay tracking based on the location reported by the monitoring device, and continuously relays the tracked person through the relay between the monitoring devices. The server side performs clustering analysis based on the captured images sent by the monitoring device and archives the images of the tracked person. In the technical solution of this application, the server dynamically determines the next-level alternative monitoring device according to the location of the monitoring device that reports the tracking loss, effectively narrowing the tracking range of the next-level alternative monitoring device, avoiding sending the information and tracking identifier of the tracked person to all monitoring devices, avoiding wasting network resources, saving monitoring device resources, and enabling the monitoring device receiving the instruction to obtain the feature information of the tracked person sent by the server as soon as possible and respond in time, effectively preventing missed tracking caused by untimely receipt of information by the monitoring device, realizing intelligent automatic continuous tracking of the same monitoring target, improving the accuracy of tracking personnel, and further improving the accuracy of archiving.

[0181] Optionally, as one of the embodiments, after determining the monitoring target, the first monitoring device and the next-level alternative monitoring device further include:

[0182] Determine the motion state of the monitoring target; the tracking loss message also includes the motion state information of the monitoring target;

[0183] The relay tracking instruction also includes: the estimated arrival time information for reaching the next-level alternative monitoring device; the estimated arrival time information is determined according to the geographical distance between the corresponding next-level alternative monitoring device and the monitoring device that sends the tracking loss message, and the motion state information of the monitoring target;

[0184] The server 201 side determines the second similarity threshold according to the estimated arrival time information and the threshold adjustment information. This step can also be performed by the next-level alternative monitoring device. In this embodiment, the following-level alternative device is used to perform the current step for illustration, and the next-level alternative monitoring device is further used to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold.

[0185] The following is illustrated by specific embodiments.

[0186] As Figure 6 shown, Figure 6 it is a flowchart of the tracking method at the monitoring device end, which is one of the next-level alternative monitoring devices.

[0187] S601: Receive the relay tracking instruction sent by the server 201. The relay tracking instruction at least includes the feature information of the person being tracked, the tracking identifier, and the estimated arrival time information at this alternative monitoring device.

[0188] The feature information of the person being tracked, the tracking identifier, and the estimated arrival time information at this alternative monitoring device belong to a kind of data information.

[0189] S602: Search for the target in the local monitoring video.

[0190] S603: Determine whether there is a monitoring target whose similarity meets the second similarity threshold.

[0191] If so, execute step S604; if not, continue to execute step S602.

[0192] S604: Periodically obtain the captured images of the monitoring target.

[0193] S605: The monitoring device sends to the server 201: Report each captured image, the corresponding capture time, and the tracking identifier to the server 201.

[0194] This process is a separate process and does not participate in the nested loop of the tracking process. That is, after the tracking ends, the server 201 can still receive information such as the captured images, the corresponding capture times, and the tracking identifiers sent by the monitoring device.

[0195] S606: Determine the motion state of the monitoring target.

[0196] S607: The monitoring device sends to the server 201: After the monitoring target disappears from the video screen, send a tracking loss message containing at least the tracking identifier to the server 201. The tracking loss message also includes the motion state information of the monitoring target.

[0197] Figure 7 It is a flowchart of the tracking method at the server end;

[0198] S701: Receive the tracking loss message sent by the monitoring device.

[0199] This message is Figure 6 the tracking loss message sent in step S607 in , and the tracking loss message at least includes the feature information of the person being tracked, the tracking identifier, and the motion state information of the monitoring target.

[0200] S702: Determine whether the tracking end condition is satisfied;

[0201] If so, execute step S704; if not, execute step S703;

[0202] S703: Based on the geographical location of the monitoring device that sent the tracking lost message, determine the monitoring devices within a preset distance range as the next-level alternative monitoring devices;

[0203] S705: The server 201 sends to the next-level alternative monitoring devices: send a relay tracking instruction to the next-level alternative monitoring devices. The relay tracking instruction includes at least the feature information and tracking identifier of the person being tracked; the relay tracking instruction also includes: the estimated arrival time information to this next-level alternative monitoring device; the estimated arrival time information is determined according to the geographical distance between the corresponding next-level alternative monitoring device and the monitoring device that sent the tracking lost message, and the motion state information of the monitoring target.

[0204] S706: Perform clustering analysis based on the captured images sent by the monitoring devices, and archive the captured images determined to belong to the person being tracked.

[0205] This process is an independent process and does not participate in the nested loop of the tracking process. The received image information is Figure 6 the captured images, corresponding capture times, tracking identifiers, and other information sent in step S605.

[0206] In the embodiment of the present invention, the server can reasonably regulate and arrange the monitoring devices for relay tracking, and adjust the similarity threshold of different monitoring devices within different estimated arrival times according to the estimated arrival time of the person being tracked at different monitoring devices, effectively improving the tracking success rate.

[0207] Optionally, as one of the embodiments, after the next-level alternative monitoring device determines that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold, it further includes: sending a tracking success message containing the tracking identifier to the server 201;

[0208] After receiving the tracking success message, the network-side device, i.e., the server 201, sends a monitoring stop message containing the tracking identifier to each of the next-level alternative monitoring devices among the most recently determined next-level alternative monitoring devices except the monitoring device that currently sent the tracking success message;

[0209] The next-level alternative monitoring device that receives the monitoring stop message stops detecting whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold.

[0210] The following is illustrated by specific embodiments.

[0211] See the appendix Figure 9 as shown. The appendix Figure 9 Compared with the appendix Figure 6 the following steps are added:

[0212] S902: Whether to receive the tracking identifier monitoring stop message sent by the server 201;

[0213] If yes, execute step S904; if no, execute step S903.

[0214] S903: Stop detecting whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold.

[0215] If the monitoring device receives the tracking identifier monitoring stop message sent by the server, it indicates that another monitoring device has detected that the tracked person appears in another monitoring device. The current monitoring device does not need to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold, and the motion state information and the estimated arrival time received by the current monitoring device from the server need to be recalculated according to the location and motion state information of the monitoring device where the tracked person arrives again. Therefore, the monitoring device needs to wait for the server to send a relay tracking instruction again, that is, step S901.

[0216] S906: Send a tracking success message containing the tracking identifier to the server 201;

[0217] When it is found that there is a monitoring target in the local monitoring video stream that meets the preset conditions, it indicates that the tracked person appears in the monitoring area of the current monitoring device, and the monitoring device has successfully tracked the tracked person. Therefore, the monitoring device sends a tracking success message containing the tracking identifier to the server to report that it has successfully tracked the tracked person, so that the server can perform the next operation.

[0218] See the appendix Figure 10 as shown. The appendix Figure 10 Compared with the appendix Figure 7 the following steps are added:

[0219] S1006: Receive the tracking success message sent by the monitoring device;

[0220] S1007: Send a monitoring stop message containing the tracking identifier to each of the next-level alternative monitoring devices determined most recently, except for the monitoring device that currently sends the tracking success message, among the next-level alternative monitoring devices;

[0221] When the server receives the tracking success message sent by the monitoring device, the same tracked person will not appear in other monitoring devices at the same time. Therefore, other monitoring devices located in the same alternative monitoring device do not need to search for the current tracked person in the local video stream, and the estimated arrival time in the relay tracking instruction sent to the next-level alternative monitoring device previously needs to be recalculated according to the time and motion state sent by the alternative monitoring device where the current tracked person arrives again. Therefore, the server sends a monitoring stop message containing the tracking identifier to each next-level alternative monitoring device except the monitoring device that currently sends the tracking success message.

[0222] In the above embodiment Figure 8 Take the embodiment as an example to supplement and explain this embodiment again.

[0223] Suppose the monitoring device 122 has monitored the tracked person, and the monitoring device 122 sends a tracking success message containing the tracking identifier to the server 201. After receiving the tracking success message, the server 201 sends a monitoring stop message containing the tracking identifier to each next-level alternative monitoring device in the most recently determined next-level alternative monitoring devices, that is, the monitoring device 122 and the monitoring device 123, except the monitoring device that currently sends the tracking success message. That is, each next-level alternative monitoring device except the monitoring device 122 that currently sends the tracking success message sends a monitoring stop message containing the tracking identifier. For example, a monitoring stop message containing the tracking identifier is sent to the monitoring device 123. If there are other monitoring devices in this alternative monitoring device, such as the monitoring device 124, it will also receive the monitoring stop message and stop monitoring the current tracked person. Other monitoring devices need to wait for the server to send a relay tracking instruction again to recalculate the estimated arrival time of the tracked person.

[0224] In the embodiment of the present invention, when the monitoring device has monitored the tracked person, it sends a tracking success message to the server. After receiving the tracking success message, the server sends a message to stop tracking the current tracked person to other monitoring devices, and other monitoring devices stop monitoring the current tracked person and wait for the server to send a relay tracking instruction again. The technical solution of the embodiment of the present invention reduces the number of unnecessary recognition feature matches of the monitoring device, saves the resources and computing power of the monitoring device, and thus achieves the efficient utilization of the monitoring device.

[0225] Optionally, as one of the embodiments, the monitoring devices within a preset distance range can be determined as the next-level alternative monitoring devices according to the geographical location of the monitoring device that sends the tracking loss message, specifically including:

[0226] According to the determined motion state of the monitoring target, select the monitoring devices within the corresponding preset distance range of the motion state and determine them as the next-level alternative monitoring devices.

[0227] The following is illustrated by specific embodiments.

[0228] Specifically, see Table 5: Mapping Table of Motion State and Preset Distance.

[0229] Table 5 Mapping Table of Motion State and Preset Distance

[0230]

[0231]

[0232] The following is through Figure 8 the following embodiments to illustrate.

[0233] If the monitoring device 121 monitors that the tracked person is in a walking state, the server searches the mapping table of motion state and preset distance according to the motion state of the tracked person. As can be seen from Table 1, the preset distance in the walking state is 500m. Then, the monitoring device with the monitoring device 121 as the center and 500m as the radius is used as the next-level alternative monitoring device. If it is monitored that the tracked person is in a running state, according to the mapping table, the preset distance in the running state is 1000m. Then, the monitoring device with the monitoring device 121 as the center and 1000m as the radius is used as the next-level alternative monitoring device.

[0234] The embodiments of the present invention determine the motion state of the monitoring target, and select the monitoring devices within the preset distance range corresponding to the motion state as the next-level alternative monitoring devices; the technical solution of the embodiments of the present invention defines different ranges of the next-level alternative monitoring devices according to different motion states of the monitored person, further narrows the range of the next-level alternative monitoring devices, relatively reduces the number of alternative monitoring devices, saves device resources, relatively speeds up the matching speed of the monitoring device for relay tracking to identify the person to be tracked, and provides a technical solution for how to select the monitoring device for relay tracking.

[0235] Optionally, as one of the embodiments, the motion state includes the motion direction; among the monitoring devices within the preset distance range corresponding to the motion state, the monitoring devices that match the motion direction are further determined as the next-level alternative monitoring devices.

[0236] The following is illustrated by specific embodiments.

[0237] See Appendix Figure 11 , the monitoring devices 121, 122, 123 and the monitoring device 125 are all located on the viaduct where only motor vehicles can pass. The direction of the connecting line is the route where motor vehicles can travel. The monitoring devices do not communicate or interact with each other, and only perform data interaction with the server 201.

[0238] The distance between the monitoring device 121 and the monitoring device 122 is 100 meters. The distance between the monitoring device 123 and the monitoring device 121 is 200 meters. The spatial distance between the monitoring device 125 and the monitoring device 121 is 100m. However, the monitoring device 125 and the monitoring device 121 are located on viaducts at different positions, and there is no passable route between them.

[0239] When the monitoring device 121 monitors that the tracked person appears in the monitoring area of the monitoring device 121, it is determined that the tracked person is in the state of driving a motor vehicle. According to the mapping table of motion state and preset distance in Table 1, all monitoring devices within a distance of 6000m are used as the next-level alternative monitoring devices. On the premise that the monitoring device 122, the monitoring device 123, and the monitoring device 125 all meet the preset distance, according to the monitoring devices matched by the motion direction, the server 201 uses the monitoring device 122 and the monitoring device 123 as the next-level alternative monitoring devices, and will not include the monitoring device 125 in the next-level alternative monitoring devices. Because according to the motion direction, it is impossible for the tracked person to drive a motor vehicle to reach the monitoring device 125. Even if the distance condition is met, considering the matching of the motion direction, the monitoring device 125 will not be used as the next-level alternative monitoring device.

[0240] At this time, in the system, the distance between the monitoring device 121 and the monitoring device 125 can be set to an infinite distance, or an exclusion method, or other methods can be adopted. When the monitoring device 121 monitors the tracked person, when selecting the next-level alternative monitoring device, the monitoring device 125 is not used as the next-level alternative monitoring device. In the actual application process, those skilled in the art can adjust according to the actual situation, and no limitation is made here.

[0241] In the embodiment of the present invention, among the monitoring devices within the preset distance range corresponding to the motion state, the monitoring devices matched by the motion direction are further determined as the next-level alternative monitoring devices, and the monitoring devices that do not conform to the traveling direction logic are excluded and not included in the monitoring devices for relay tracking, further narrowing the range of monitoring devices for relay tracking, avoiding interference to the tracking monitoring devices caused by sending meaningless tracking instructions, and accelerating the matching speed of the monitoring devices for relay tracking to identify the person to be tracked.

[0242] Optionally, as one of the embodiments, after the network-side device determines that the tracking end condition is met, it sends a tracking end instruction to the next-level alternative monitoring device determined most recently; the next-level alternative monitoring device that receives the tracking end instruction stops detecting whether there is a monitoring target whose similarity meets the second similarity threshold in the local monitoring video stream, and stops sending tracking loss messages and tracking success messages to the network-side device.

[0243] The following is illustrated by specific embodiments.

[0244] See the appendix Figure 5 , and the other steps are the same as those in the above embodiments. The content that is the same as the above embodiments will not be elaborated here, and only the differences from the above embodiments will be described.

[0245] S54042: The server 201 determines whether the tracking end condition is satisfied;

[0246] If not, execute step S54043; if so, execute step S5601;

[0247] S5601: The server sends a tracking end instruction to the next-level alternative monitoring device determined most recently;

[0248] S5602: The next-level alternative monitoring device that receives the tracking end instruction stops detecting whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold, and stops sending tracking loss messages and tracking success messages to the server 201.

[0249] See the appendix Figure 7 , and the other steps are the same as those in the above embodiments. The content that is the same as the above embodiments will not be elaborated here, and only the differences from the above embodiments will be described.

[0250] S702: Whether the tracking end condition is satisfied;

[0251] If so, execute step S704; if not, execute step S703.

[0252] S704: Send a tracking end instruction to the next-level alternative monitoring device determined most recently;

[0253] The tracking end condition can be defined by oneself and can be freely set or adjusted. For example, directly issue an instruction not to track the current tracked person anymore, or end the tracking when the tracked person has not been tracked for a certain period of time, or stop tracking when the set distance range is exceeded, or the characteristic information of the tracked person provided is not sufficient to match the monitoring target, etc. It can also be other end conditions. This is only an example here and is not limited. Those skilled in the art can define the tracking end condition according to actual needs.

[0254] Take Figure 8 as an example to make a supplementary explanation for this embodiment.

[0255] If the monitoring device 121 monitors the tracked person and the tracked person has left the monitoring area of the monitoring device 121, the server 201 determines the next-level alternative monitoring devices as the monitoring device 122 and the monitoring device 123 and sends a relay tracking instruction.

[0256] At this time, when the server 201 receives the condition that the tracking end is determined to be satisfied, it sends a tracking end instruction to the last determined next-level alternative monitoring device, that is, sends a tracking end instruction to the monitoring device 122 and the monitoring device 123.

[0257] After the monitoring device 122 and the monitoring device 123 receive the tracking end instruction sent by the server 201, they stop detecting whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold, and stop sending the tracking loss message and the tracking success message to the server 201. That is, for the monitoring device 122 and the monitoring device 123, regardless of whether the tracking is successful, they no longer send the information of the currently tracked person to the server 201, nor do they detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold.

[0258] In the embodiment of the present invention, by adding a determination of whether the tracking end condition is met, it avoids wasting the computing power of the monitoring device when the tracked person cannot be tracked for a long time or when there is no need to track the currently tracked person, and improves the usage efficiency of the monitoring device.

[0259] Optionally, as one of the embodiments, adopting the person tracking method in any of the above embodiments further includes: obtaining the data information of the tracked person, where the data information is data information including at least the captured image of the tracked person;

[0260] According to the data information of the tracked person, perform clustering analysis and file the captured images determined to be of the tracked person.

[0261] Performing clustering analysis and filing the captured images determined to be of the tracked person specifically includes:

[0262] At every set time interval, perform a clustering analysis on all the captured images determined to be of the tracked person and file the captured images belonging to the tracked person; or

[0263] After the time interval without receiving the data information of the tracked person exceeds the set threshold, perform a clustering analysis on all the captured images determined to be of the tracked person that have been received and file the captured images determined to be of the tracked person.

[0264] Filing the captured images determined to be of the tracked person specifically includes:

[0265] The data information further includes the capture time corresponding to the captured image;

[0266] Sort the corresponding monitoring devices that send the captured images in the chronological order of the capture times corresponding to the captured images after archiving, to obtain a virtual movement trajectory of the person being tracked, which is composed of each monitoring device as a trajectory node;

[0267] Determine unreasonable trajectory nodes according to the actual geographical location information of each monitoring device in the virtual movement trajectory;

[0268] Delete the corresponding captured images reported by the unreasonable trajectory nodes.

[0269] The following is illustrated by specific embodiments.

[0270] In the embodiment of the present application, the server 201 performs a clustering analysis on all the captured images received corresponding to the tracking identifier at intervals of a set duration, and archives the captured images determined to belong to the person being tracked. For example, a clustering analysis and archiving are performed every 30 minutes; or after the interval duration without receiving new captured images by the server 201 exceeds a set threshold, such as exceeding 24 hours, the server 201 itself performs a clustering analysis on all the captured images received corresponding to the tracking identifier, and archives the captured images determined to belong to the person being tracked.

[0271] The server 201 in this embodiment can perform a clustering analysis on the captured images by means including but not limited to partitioning clustering, hierarchical clustering, density clustering, etc. By analyzing the feature information in the captured images sent by each monitoring device, it analyzes whether it is the same person being tracked, and excludes the image information that does not belong to the same person being tracked to reduce incorrect archiving. The server 201 can perform a clustering analysis regularly, or perform a clustering analysis after not receiving captured images sent by the monitoring device for more than a certain time.

[0272] The technical solution of the embodiment of the present invention improves the real-time performance and accuracy of archiving the person being tracked through the server's regular clustering analysis and sorting of the captured images.

[0273] Another embodiment is shown in Figure 11 As shown, assume the distance between the monitoring device 121 and the monitoring device 122 is 1000m, and the spatial distance between the monitoring device 125 and the monitoring device 121 is 800m. The direction indicated by the connecting line and the arrow is the feasible traveling direction, and there is no connecting line for non-reachable routes, that is, there is no reachable route between the monitoring device 121 and the monitoring device 125.

[0274] If the server 201 receives the captured image of the tracked person reported by the monitoring device 121, and based on the time and the location information of the monitoring device 121, it determines that the tracked person is currently near the monitoring device 121. However, in the next second after the monitoring device 121 reports the captured image, the server 201 receives the image information of the captured tracked person sent by the monitoring device 125. Based on the capture time of the monitoring device 125 and the location information of the monitoring device 125, a virtual movement trajectory of the tracked person is formed. The tracked person was at the monitoring device 121 in the previous second and then quickly appeared at the monitoring device 125. There is no travel route between the monitoring device 121 and the monitoring device 125. No matter what travel mode the tracked person uses, it is impossible to appear at the monitoring device 125 in the second second. Therefore, one of the monitoring device 121 and the monitoring device 125 has an unreasonable trajectory.

[0275] However, at this time, it is impossible to determine which monitoring device's trajectory information is unreasonable. When the next monitoring device reports tracking the tracked person and uploads the captured image, the server 201 will also obtain the current information. If 60s after the monitoring device 121 captures the image of the tracked person, the server receives the captured image of the tracked person sent by the monitoring device 122, a virtual movement trajectory of the tracked person is formed according to the travel route and time.

[0276] Suppose the time when the monitoring device 121 uploads the image of the tracked person is 9:00. Then, the virtual movement trajectory of the tracked person is formed as follows:

[0277] Location: Monitoring device 121, time 9:00:00 ->

[0278] Location: Monitoring device 125, time 9:00:01 ->

[0279] Location: Monitoring device 122, time 9:01:00;

[0280] From the above virtual movement trajectory, it can be determined that the information at the monitoring device 125 is an unreasonable trajectory node. For the same tracked person, it is impossible to appear at different locations within a time difference of 1s, and there is no fast travel route between the monitoring device 121 and the monitoring device 125. Therefore, it is necessary to delete the corresponding captured image reported by the unreasonable trajectory node, that is, delete the corresponding captured image reported by the monitoring device 125. This method is to eliminate unreasonable data by determining the virtual movement trajectory of the tracked person.

[0281] The technical solution of the embodiment of the present invention forms a virtual trajectory of the person to be tracked through the capture times of multiple monitoring devices and the location information of each monitoring device, determines the rationality of the person data, eliminates the unreasonable data, and this method performs person clustering analysis by combining feature algorithms and tracking algorithms, excludes image information that does not belong to the same person to be tracked to reduce incorrect archiving, and improves the accuracy of archiving the person to be tracked.

[0282] Optionally, as one of the embodiments, a person tracking device is further provided, including:

[0283] A receiving module, configured to receive a tracking loss message sent by a first monitoring device, where the tracking loss message at least includes the feature information of the person to be tracked;

[0284] A relay tracking module, configured to determine a next-level alternative monitoring device according to the first monitoring device, and determine the threshold adjustment information of the next-level alternative monitoring device; determine a second similarity threshold according to the threshold adjustment information; wherein, the next-level alternative monitoring device is configured to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction;

[0285] A sending module, configured to send a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction at least includes the feature information of the person to be tracked and the threshold adjustment information.

[0286] Optionally, as one of the embodiments, a person archiving device is further provided. Using the person tracking device of the above embodiment, it further includes:

[0287] An obtaining module, which obtains the data information sent by the person tracking device. The data information is the data information sent by the monitoring device of the tracking device after confirming that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and confirming that the monitoring target is the person to be tracked, and at least includes the captured image of the person to be tracked and the corresponding capture time;

[0288] An archiving module, configured to perform clustering analysis according to the data information of the person to be tracked, and archive the captured images determined to be the person to be tracked.

[0289] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0290] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to the application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows 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 the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0291] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0292] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 block or multiple blocks.

[0293] Obviously, those skilled in the art can make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalent technologies, this application is also intended to include these changes and modifications.

Claims

1. A personnel tracking method, characterized in that, The method includes: Receiving a tracking loss message sent by a first monitoring device, where the tracking loss message includes at least the feature information of the person being tracked; Determining a next-level alternative monitoring device according to the first monitoring device, and determining threshold adjustment information for the next-level alternative monitoring device; Sending a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction includes at least the feature information of the person being tracked and the threshold adjustment information; Determining a second similarity threshold according to the threshold adjustment information; Wherein, the next-level alternative monitoring device is used to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction; The threshold adjustment information is determined according to at least any one of the motion state of the person being tracked obtained by the first monitoring device, the geographical distance between the second monitoring device in the next-level alternative monitoring device and the first monitoring device, and the estimated arrival time for the person being tracked to reach the second monitoring device in the next-level alternative monitoring device; The determining the second similarity threshold according to the threshold adjustment information specifically includes: Determining the second similarity threshold according to the different arrival times of the person being tracked and the threshold adjustment information; wherein, the different arrival times are the different times for the person being tracked to reach the second monitoring device in the next-level alternative monitoring device.

2. The method according to claim 1, wherein The determining the next-level alternative monitoring device according to the first monitoring device specifically includes: Determining the next-level alternative monitoring device according to the geographical location of the first monitoring device.

3. The method according to claim 2, characterized in that, The determining the next-level alternative monitoring device according to the geographical location of the first monitoring device specifically includes: Determining the monitoring devices within a preset distance range as the next-level alternative monitoring devices according to the geographical location of the first monitoring device.

4. The method according to claim 3, wherein The determining the monitoring devices within a preset distance range as the next-level alternative monitoring devices according to the geographical location of the first monitoring device specifically includes: The tracking loss message includes the motion state information of the person being tracked; Selecting the monitoring devices within a preset distance range corresponding to the motion state according to the motion state information as the next-level alternative monitoring devices.

5. The method according to claim 1, characterized in that The tracking loss message sent by the first monitoring device is specifically: The tracking loss message sent by the first monitoring device when the person being tracked does not meet the monitoring conditions of the first monitoring device.

6. The method according to claim 1, wherein: The tracking loss message further includes the motion state information of the person being tracked; The relay tracking instruction further includes: estimated arrival time information for reaching the next-level alternative monitoring device; Wherein, the estimated arrival time information is determined according to the geographical distance between the second monitoring device in the next-level alternative monitoring device and the first monitoring device, and the motion state information of the monitoring target.

7. The method according to claim 6, wherein It further includes: Determining the second similarity threshold according to the estimated arrival time information and the threshold adjustment information.

8. The method according to claim 1, characterized in that, The next-level alternative monitoring device is further used to send a tracking success message for the person being tracked; Among them, the tracking success message is sent by the next-level alternative monitoring device after determining that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and determining that the monitoring target is the person to be tracked.

9. The method according to claim 8, wherein It further includes: After receiving the tracking success message of the person to be tracked, send a monitoring stop message of the person to be tracked to each of the next-level alternative monitoring devices determined most recently, except for the monitoring device that currently sends the tracking success message, among the next-level alternative monitoring devices.

10. The method according to claim 1, wherein The receiving the tracking loss message sent by the first monitoring device specifically includes: According to the tracking end condition, confirm that when the tracking end condition is not met, receive the tracking loss message sent by the first monitoring device; and / or When the tracking end condition is met, send a tracking end instruction to the next-level alternative monitoring device determined most recently and do not receive the tracking loss message sent by the monitoring device.

11. A personnel filing method, characterized in that, Using the person tracking method according to any one of claims 1-10, it further includes: Obtain the data information of the person to be tracked, where the data information is sent by the monitoring device after determining that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and determining that the monitoring target is the person to be tracked, and at least includes the captured image of the person to be tracked and the data information corresponding to the capture time; According to the data information of the person to be tracked, perform clustering analysis and file the captured images determined to be of the person to be tracked.

12. The method according to claim 11, wherein The performing clustering analysis and filing the captured images determined to be of the person to be tracked specifically includes: Every set time interval, perform a clustering analysis on all the captured images determined to be of the person to be tracked and file the captured images belonging to the person to be tracked; or After the time interval without receiving the data information of the person to be tracked exceeds the set threshold, perform clustering analysis on all the captured images determined to be of the person to be tracked that have been received and file the captured images determined to be of the person to be tracked.

13. The method according to claim 12, characterized in that, The filing the captured images determined to be of the person to be tracked specifically includes: Sort the corresponding monitoring devices that send the corresponding captured images in the order of the capture time corresponding to the filed captured images to obtain a virtual movement trajectory of the person to be tracked composed of each monitoring device as a trajectory node; According to the actual geographical location information of each monitoring device in the virtual movement trajectory, determine the unreasonable trajectory nodes; Delete the corresponding captured images reported by the unreasonable trajectory nodes.

14. A personnel tracking device, characterized in that, It includes: A receiving module, configured to receive the tracking loss message sent by the first monitoring device, where the tracking loss message at least includes the feature information of the person to be tracked; A relay tracking module, configured to determine the next-level alternative monitoring device according to the first monitoring device and determine the threshold adjustment information of the next-level alternative monitoring device; determine the second similarity threshold according to the threshold adjustment information; where the next-level alternative monitoring device is configured to detect whether there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold according to the received relay tracking instruction; A sending module, configured to send a relay tracking instruction to the next-level alternative monitoring device, where the relay tracking instruction at least includes feature information of the person to be tracked and the threshold adjustment information; Wherein, the threshold adjustment information is determined according to at least any one of the motion state of the person to be tracked obtained by the first monitoring device, the geographical distance between the second monitoring device in the next-level alternative monitoring device and the first monitoring device, and the preset arrival time of the person to be tracked at the second monitoring device in the next-level alternative monitoring device; Determining the second similarity threshold according to the threshold adjustment information specifically includes: Determining the second similarity threshold according to different arrival times of the person to be tracked and the threshold adjustment information; wherein, the different arrival times are different times when the person to be tracked arrives at the second monitoring device in the next-level alternative monitoring device.

15. A personnel filing device, characterized in that, Using the personnel tracking device as described in claim 14, further comprising: An acquisition module, configured to acquire data information sent by the personnel tracking device, where the data information is data information sent by a monitoring device of the tracking device after confirming that there is a monitoring target in the local monitoring video stream whose similarity meets the second similarity threshold and confirming that the monitoring target is the person to be tracked, and at least includes a captured image of the person to be tracked and the corresponding capture time; An archiving module, configured to perform clustering analysis according to the data information of the person to be tracked and archive the captured images determined to be of the person to be tracked.

Images