A method and system for tracking suspicious persons based on companions based on time-space trajectory
Through the peer trajectory analysis model, the problem of missing trajectory of suspicious personnel caused by light, angle or occlusion in traditional technology is solved, and the completion and tracking of the trajectory of suspicious personnel is achieved, and the case handling efficiency is improved.
Patent Information
- Application Number
- CN202111134553.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-09-27
- Publication Date
- 2024-12-31
- Estimated Expiration
- 2041-09-27
AI Technical Summary
Traditional suspicious person tracking technology can easily lead to the loss of targets under light, angle or intentional occlusion, and the trajectory routes of suspicious persons cannot be accurately identified and recorded, resulting in increased tracking difficulties and manual investigations.
Through the peer trajectory analysis model, peers who appear at the same time with suspicious people within the set time are selected, their time and space trajectory information are analyzed, face clustering and similarity comparison are performed, and the trajectory information of suspicious people is completed, so as to achieve a long time track.
It effectively reduces the manual investigation process, quickly finds and locates suspicious people, improves case handling efficiency, and ensures the integrity of the trajectory of suspicious people.
Smart Images
Figure CN113934889B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of target tracking, and in particular to a method and system for tracking suspicious persons according to their companions based on time-space trajectory. Background Art
[0002] Face recognition is a biometric technology that identifies people based on their facial features. It uses a camera or a camera to collect images or video streams containing faces, and automatically detects and tracks faces in the images, and then performs a series of related technologies on the detected faces, which are usually called portrait recognition or facial recognition. In recent years, face recognition technology has been widely used in the security field. With the continuous development of technology, intelligent detection technologies such as precise positioning, target recognition and specific target tracking based on surveillance video equipment have become an important means of detection to maintain national security and social stability under the new situation.
[0003] Traditional suspicious person tracking technology automatically identifies and tracks people by installing surveillance video equipment at various intersections, shopping malls, etc. to collect faces. However, under the monitoring perspective, the accuracy of face recognition will decrease when the suspicious person is walking due to poor lighting and angles; if the suspicious person intentionally uses a hat or an umbrella to cover his face, it is impossible to use face recognition to track the movement trajectory of the suspicious person for a long time. Therefore, technicians in this field are committed to developing a method and system for tracking suspicious persons based on companions based on spatiotemporal trajectories. When the suspicious person loses the tracking target due to light, angle or deliberate cover, and cannot accurately identify and record his trajectory, the suspicious person can be tracked from the trajectory of the companions to complete the trajectory of the suspicious person, thereby assisting the police in tracking and reducing the manual investigation process. Summary of the invention
[0004] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is how to solve the problem of losing the target of a suspicious person due to light, angle or deliberate obstruction during long-term tracking of a suspicious person. When the track of a suspicious person is lost, the analysis model of the track of the suspicious person is analyzed by the companions to track the suspicious person from the track of the companions, and the track route of the suspicious person can be completed.
[0005] To achieve the above object, the present invention provides a method for tracking a suspicious person based on a companion based on a spatiotemporal trajectory, comprising the following steps:
[0006] Step 1: Screen out suspicious persons and their companions from their photos within a set range and time;
[0007] Step 2: Analyze and save the space-time trajectory information of the companions at all checkpoints within the set range and time according to the companions' photos, and save it as the companions' space-time trajectory information;
[0008] Step 3: Based on the spatiotemporal trajectory information of fellow travelers, photos of all pedestrians that appear within a specified time threshold before and after the same time point at each checkpoint are screened out, and face clustering is performed. The clustered photos and corresponding spatiotemporal trajectory information are saved in a list of suspected suspicious persons;
[0009] Step 4: Screen and identify suspicious persons by matching the photos in the list of suspected suspicious persons with the photos of the suspicious persons;
[0010] Step 5: Use the space-time trajectory information in the corresponding list of suspected suspicious persons to complete the space-time trajectory information of the suspicious person and continue tracking.
[0011] Furthermore, step 1 includes the following steps:
[0012] Step 1.1: Use the monitoring equipment at each checkpoint to capture photos of the suspicious person and his companions within the set range and time;
[0013] Step 1.2: Confirm the space-time trajectory information of the suspicious person through the monitoring equipment at each checkpoint, analyze the space-time trajectory information of the suspicious person and the space-time trajectory information of the companions, and when the overlap between the two reaches a first set threshold, save the screened persons as companions.
[0014] Furthermore, each checkpoint monitoring device includes multiple snapshot cameras that capture photos simultaneously, and selects photos whose facial image information is closest to the standard facial image information.
[0015] Further, step 4 includes the following steps:
[0016] Step 4.1, comparing the photos in the suspected suspicious person list with the photos of the suspicious person, and marking the photos whose similarity exceeds a second set threshold;
[0017] Step 4.2: Then manually screen the marked photos to identify suspicious persons.
[0018] Further, in step 3, the specified time threshold is set to 10 seconds.
[0019] The present invention also provides a system for tracking suspicious persons based on companions based on spatiotemporal trajectory, comprising a screening module, a companion trajectory storage module, a companion trajectory analysis module, a suspicious person judgment module and a trajectory information storage module, wherein:
[0020] A screening module is used to screen out suspicious persons and their companions from photos of them within a set range and a set time;
[0021] The companion trajectory storage module is used to analyze and save the space-time trajectory information of the companions appearing at all checkpoints within a set range and a set time according to the companions' photos, and save it as the companions' space-time trajectory information;
[0022] The companion trajectory analysis module is used to filter out the photos of all pedestrians that appear within a specified time threshold before and after the same time point at each checkpoint based on the companion's spatiotemporal trajectory information, and perform face clustering, and save the clustered photos and corresponding spatiotemporal trajectory information in the list of suspected suspicious persons;
[0023] A suspicious person identification module is used to screen out suspicious persons by matching the photos in the list of suspected suspicious persons with the photos of the suspicious persons;
[0024] The trajectory information storage module is used to complete the spatiotemporal trajectory information of suspicious persons and continue tracking them.
[0025] Furthermore, the screening module utilizes the monitoring equipment at each checkpoint to capture photos of suspicious persons and companions within a set range and a set time, and confirms the spatiotemporal trajectory information of the suspicious persons through the monitoring equipment at each checkpoint, and analyzes the spatiotemporal trajectory information of the suspicious persons and the spatiotemporal trajectory information of the companions. When the overlap between the two reaches a first set threshold, the screened persons are saved as companions.
[0026] Furthermore, the suspicious person judgment module compares the photos in the suspected suspicious person list with the photos of the suspicious person for similarity, and marks the photos whose similarity exceeds a second set threshold; then manually screens the marked photos to find the suspicious person.
[0027] Furthermore, in the companion trajectory analysis module, the specified time threshold is set to 10 seconds.
[0028] Furthermore, in the screening module, each checkpoint monitoring device includes multiple snapshot cameras that simultaneously capture photos, and screen out photos whose facial image information is closest to the standard facial image information.
[0029] Compared with the prior art solutions, the beneficial effects brought by the solution of the present invention are:
[0030] The present invention tracks the trajectory of suspicious persons through an analysis model of the trajectory of suspicious persons analyzed by their companions. When the tracking target is lost and cannot be accurately identified due to light, angle or deliberate obstruction, the spatial-temporal trajectory information of the companions can be combined to screen the persons who appear at the same time as the companions within a specified time range through facial feature clustering and similarity comparison, thereby reducing the process of manual screening of suspicious persons, facilitating rapid search and positioning of suspicious persons, and completing the spatial-temporal trajectory route of the suspicious persons, thereby achieving the purpose of long-term tracking of suspicious persons and effectively improving case handling efficiency.
[0031] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 It is a flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION
[0033] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.
[0034] In the drawings, components with the same structure are indicated by the same numerical reference numerals, and components with similar structures or functions are indicated by similar numerical reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the components is appropriately exaggerated in some places in the drawings.
[0035] like Figure 1 As shown, in a preferred embodiment of the present invention, within a set range and a set time, multiple snapshot cameras of each checkpoint monitoring device simultaneously capture photos, and select photos whose facial image information is closest to the standard facial image information to obtain the trajectory information of the suspicious person A, which is stored in Table 1. Table 1 is as follows:
[0036]
[0037]
[0038] Table 1
[0039] According to the companions of the suspicious person A, the companion B is selected, and according to the photo of the companion B, the spatiotemporal trajectory information of the companion B appearing at all checkpoints within the set range and the set time is analyzed and saved, and saved as the spatiotemporal trajectory information of the companion, and stored in Table 2. Table 2 is as follows:
[0040] Serial number time Place 1 10:30 Bayonet 1 2 10:35 Bayonet 2 3 10:36 Bayonet 3 4 10:48 Bayonet 4 5 10:52 Bayonet 5 6 11:28 Bayonet 6 7 11:37 Bayonet 7 8 12:17 Bayonet 8 9 13:11 Bayonet 9 10 13:54 Bayonet 10 11 14:29 Bayonet 11 12 15:30 Bayonet 12
[0041] Table 2
[0042] According to the spatiotemporal trajectory information of companion B, people who appear within a specified 10-second difference before and after the same time point at each checkpoint are screened out, and face clustering is performed. The clustered photos and the corresponding spatiotemporal trajectory information are saved in a list of suspected suspicious persons. The list of suspected suspicious persons includes suspected suspicious persons C, suspected suspicious person D, suspected suspicious person E, and suspected suspicious person F.
[0043] The photos of suspected suspicious persons C, suspected suspicious persons D, suspected suspicious persons E, and suspected suspicious persons F are respectively compared with the photo of suspected person A for similarity, and suspected suspicious persons C and suspected suspicious persons D whose similarity exceeds 50% are marked.
[0044] Personnel to be analyzed Similarity C 60％ D 51％ E 15％ F 11％
[0045] Table 3
[0046] Further, through manual judgment, suspicious person A is found. Since the space-time trajectory of suspicious person A is the same as that of companion B, the space-time trajectory information of suspicious person A can be supplemented according to the space-time trajectory information of companion B, and the tracking can be continued.
[0047] Corresponding to the aforementioned embodiment of the method for tracking suspicious persons based on companions based on spatiotemporal trajectory, the present application also provides an embodiment of a system for tracking suspicious persons based on companions based on spatiotemporal trajectory. The embodiment of the system for tracking suspicious persons based on companions based on spatiotemporal trajectory of the present application can be applied to back-end devices such as platforms, and can be implemented by software, hardware, or a combination of software and hardware.
[0048] The system specifically includes: a screening module, a companion trajectory storage module, a companion trajectory analysis module, a suspicious person judgment module and a trajectory information storage module, wherein:
[0049] The screening module is used to screen out suspicious persons and their companions from the photos of suspicious persons and their companions within a set range and a set time; wherein the screening module uses each checkpoint monitoring device to capture photos of suspicious persons and their companions within a set range and a set time, and confirms the spatiotemporal trajectory information of the suspicious persons through each checkpoint monitoring device, analyzes the spatiotemporal trajectory information of the suspicious persons and the spatiotemporal trajectory information of the companions, and when the overlap between the two reaches a first set threshold, the screened persons are saved as companions. In the screening module, each checkpoint monitoring device includes multiple snapshot cameras that capture photos at the same time, and selects photos whose facial image information is closest to the standard facial image information.
[0050] The companion trajectory storage module is used to analyze and save the spatiotemporal trajectory information of companions appearing at all checkpoints within a set range and a set time based on the companions' photos, and save it as the companions' spatiotemporal trajectory information; among them, in the companion trajectory analysis module, the specified time threshold is set to 10 seconds.
[0051] The companion trajectory analysis module is used to filter out photos of all pedestrians that appear within a specified time threshold before and after the same time point at each checkpoint based on the companion's spatiotemporal trajectory information, and perform face clustering, and save the clustered photos and corresponding spatiotemporal trajectory information in a list of suspected suspicious persons.
[0052] The suspicious person judgment module is used to screen out suspicious persons by matching the photos in the suspected suspicious person list with the photos of the suspicious persons; wherein the suspicious person judgment module compares the similarity between the photos in the suspected suspicious person list and the photos of the suspicious persons, and marks the photos whose similarity exceeds a second set threshold; and then manually screens the marked photos to find suspicious persons.
[0053] The trajectory information storage module is used to complete the spatiotemporal trajectory information of suspicious persons and continue tracking them.
[0054] The implementation process of the functions and effects of each module in the above system is specifically described in the implementation process of the corresponding steps in the above method, which will not be repeated here.
[0055] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.
Claims
1. A method for tracking suspicious persons based on their companions based on spatiotemporal trajectories, characterized in that: The method comprises the following steps: Step 1: Screen out suspicious persons and their companions from their photos within a set range and time; Step 2: Analyze and save the space-time trajectory information of the companion at all checkpoints within a set range and a set time according to the companion's photo, and save it as the companion's space-time trajectory information; Step 3: Based on the spatiotemporal trajectory information of the companions, photos of all pedestrians that appear within a specified time threshold before and after the same time point at each checkpoint are screened out, and face clustering is performed, and the clustered photos and corresponding spatiotemporal trajectory information are saved in a list of suspected suspicious persons; Step 4: Screen and find the suspicious person by matching the photo in the list of suspected suspicious persons with the photo of the suspicious person; Step 5: Use the corresponding spatiotemporal trajectory information in the list of suspected suspicious persons to complete the spatiotemporal trajectory information of the suspicious person and continue tracking.
2. The method for tracking a suspicious person based on a companion based on a spatiotemporal trajectory as claimed in claim 1, characterized in that: The step 1 comprises the following steps: Step 1.1, using each checkpoint monitoring device to capture photos of the suspicious person and the traveling companions within a set range and a set time; Step 1.2: confirm the spatiotemporal trajectory information of the suspicious person through the monitoring devices at each checkpoint, analyze the spatiotemporal trajectory information of the suspicious person and the spatiotemporal trajectory information of the companions, and when the overlap between the two reaches a first set threshold, save the screened persons as the companions.
3. The method for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 2, characterized in that: Each checkpoint monitoring device includes multiple snapshot cameras that capture photos simultaneously, and selects photos whose facial image information is closest to the standard facial image information.
4. The method for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 1, characterized in that: The step 4 comprises the following steps: Step 4.1, comparing the photos in the suspected suspicious person list with the photos of the suspicious person for similarity, and marking the photos whose similarity exceeds a second set threshold; Step 4.2: Then manually screen the marked photos to find the suspicious persons.
5. The method for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 1, characterized in that: In step 3, the specified time threshold is set to 10 seconds.
6. A system for tracking suspicious persons based on their companions based on time-space trajectory, characterized in that: It includes a screening module, a companion trajectory storage module, a companion trajectory analysis module, a suspicious person judgment module and a trajectory information storage module, among which: A screening module is used to screen out suspicious persons and their companions from photos of them within a set range and a set time; A companion trajectory storage module is used to analyze and save the spatiotemporal trajectory information of the companion appearing at all checkpoints within a set range and a set time according to the companion's photo, and save it as the companion's spatiotemporal trajectory information; The companion trajectory analysis module is used to filter out the photos of all pedestrians that appear within a specified time threshold before and after the same time point at each checkpoint based on the companion spatiotemporal trajectory information, perform face clustering, and save the clustered photos and corresponding spatiotemporal trajectory information in a list of suspected suspicious persons; A suspicious person judging module, used for screening and finding the suspicious person by matching the photo in the suspected suspicious person list with the photo of the suspicious person; The trajectory information storage module is used to complete the spatiotemporal trajectory information of the suspicious person and continue tracking.
7. The system for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 6, characterized in that: The screening module utilizes each checkpoint monitoring device to capture photos of the suspicious person and the companions within a set range and a set time, and confirms the space-time trajectory information of the suspicious person through the each checkpoint monitoring device, analyzes the space-time trajectory information of the suspicious person and the space-time trajectory information of the companions, and when the overlap between the two reaches a first set threshold, the screened persons are saved as the companions.
8. The system for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 6, characterized in that: The suspicious person judgment module compares the photos in the suspected suspicious person list with the photos of the suspicious person for similarity, and marks the photos whose similarity exceeds a second set threshold; then manually screens the marked photos to find the suspicious person.
9. The system for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 6, characterized in that: In the companion trajectory analysis module, the specified time threshold is set to 10 seconds.
10. The system for tracking suspicious persons based on their companions based on time-space trajectory as claimed in claim 7, characterized in that: In the screening module, each checkpoint monitoring device includes multiple snapshot cameras that simultaneously capture photos, and screens out photos whose facial image information is closest to the standard facial image information.