Search

CN-116883899-B - Target tracking method, target tracking system, storage medium and electronic equipment

CN116883899BCN 116883899 BCN116883899 BCN 116883899BCN-116883899-B

Abstract

The application provides a target tracking method, a target tracking system, a storage medium and electronic equipment, which relate to the field of image processing, wherein the method comprises the steps of obtaining video stream data; detecting whether an occluded target exists in video stream data, if so, judging whether the occluded target can be effectively detected, if so, judging whether the upper boundary of a target detection frame of the occluded target is displaced, if so, keeping tracking the occluded target, and if not, executing a frame locking operation from the current frame to track the occluded target, wherein the frame locking operation is used for fixing the size and the position of a tracking frame. The application can avoid the tracking ID of the target in the process of switching tracking, inhibit multiple beats, improve missing beats and effectively improve the accuracy of target tracking.

Inventors

  • Duan Peicong

Assignees

  • 济南博观智能科技有限公司

Dates

Publication Date
20260505
Application Date
20230712

Claims (10)

  1. 1. A target tracking method, comprising: Acquiring video stream data; detecting whether an occluded target exists in the video stream data; If yes, judging whether the blocked target can be effectively detected; if the blocked target is effectively detected, judging whether the upper boundary of a target detection frame of the blocked target is displaced or not; if the upper boundary is displaced, keeping track of the blocked target; And if the upper boundary is not displaced and the lower boundary is moved, performing a frame locking operation from the current frame to track the blocked target, wherein the frame locking operation is used for fixing the size and the position of a tracking frame.
  2. 2. The method of claim 1, wherein detecting whether an occluded object is present in the video stream data comprises: judging whether video frames exist in the video stream data, wherein the intersection ratio between a target detection frame and a suspected shielding object detection frame is larger than zero, and the vertical axis coordinate of the lower boundary of the target detection frame is smaller than that of the lower boundary of the suspected shielding object detection frame; If yes, confirming that the blocked target exists in the video stream data.
  3. 3. The object tracking method as claimed in claim 1, wherein determining whether an upper boundary of an object detection frame of the blocked object is displaced comprises: Determining a starting frame number of the blocked target in a static state; Determining the center point coordinates of the target detection frame in the video frame corresponding to the initial frame number; If the difference value between the center point coordinate and the second center point coordinate of the target detection frame in the next video frame is smaller than a set threshold value, determining that the upper boundary of the target detection frame of the blocked target is not displaced; And if the difference value between the center point coordinate and the second center point coordinate is not smaller than the set threshold value, determining that the upper boundary of the target detection frame of the blocked target is displaced.
  4. 4. The method according to claim 1, wherein after the determining whether the blocked target can be effectively detected, further comprising: if the blocked target cannot be detected, judging whether the blocked target is contained in a low-resolution detection frame with the confidence coefficient lower than a preset value; if the blocked target is contained, continuing to track the blocked target; And if the low-resolution detection frame does not contain the blocked target, executing the frame locking operation from the current frame.
  5. 5. The object tracking method according to claim 4, wherein before determining whether the blocked object is included in the low-resolution detection frame having a confidence level lower than a preset value, further comprising: and performing target detection on the video stream data, taking a detection frame with the confidence coefficient larger than the preset value as a high-score detection frame, and taking a detection frame with the confidence coefficient lower than the preset value as a low-score detection frame.
  6. 6. The object tracking method according to claim 5, further comprising, after acquiring the video stream data: Recording the position information of the center point of the detection frame corresponding to the target; Fitting according to the position information of the central point to obtain a detection frame track; track prediction is carried out on the target by utilizing a tracking prediction algorithm to obtain a predicted frame track; determining an included angle between the detection frame track and the prediction frame track; and constructing a cost function based on the included angle, wherein the cost function is used as a matching standard between the detection frame and the tracking track.
  7. 7. The method of claim 6, wherein after fitting according to the center point position information to obtain a detection frame track, further comprising: matching the high-resolution detection frame with the tracking track of each target; For the tracking track which is not matched with the high-resolution detection frame, after the cost threshold value of the cost function is reduced, the step of matching the tracking track with the high-resolution detection frame is continuously executed; If there is an unmatched tracking track after the step of reducing the cost function threshold in the matching process and continuously executing the matching with the tracking track by using the high-resolution detection frame, matching with the unmatched tracking track by using the low-resolution detection frame; If an unknown tracking track which cannot be matched with the high-resolution detection frame and the low-resolution detection frame exists, performing frame locking operation on a preset number of video frames, and matching when a target corresponding to the unknown tracking track appears for the second time; deleting the unknown tracking track if no matchable target exists after the preset number of video frames; If an unknown high-resolution detection frame which is not matched with the tracking track exists, a tracking track is newly established, and a target corresponding to the unknown high-resolution detection frame is used as a new target.
  8. 8. A target tracking system, comprising: The video data acquisition module is used for acquiring video stream data; the shielding judging module is used for detecting whether a shielded target exists in the video stream data; The shielding detection module is used for judging whether the shielded target can be effectively detected when the judgment result of the shielding judgment module is yes; The displacement detection module is used for judging whether the upper boundary of the target detection frame of the shielded target is displaced or not when the judgment result of the shielding detection module is yes; The device comprises a tracking module, a frame locking operation and a frame locking operation, wherein the tracking module is used for keeping tracking of the blocked target if the upper boundary is displaced, and for executing the frame locking operation from the current frame to track the blocked target if the upper boundary is not displaced and the lower boundary is moved, and the frame locking operation is used for fixing the size and the position of a tracking frame.
  9. 9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the object tracking method according to any of claims 1-7.
  10. 10. An electronic device comprising a memory in which a computer program is stored and a processor that when invoked performs the steps of the object tracking method of any one of claims 1-7.

Description

Target tracking method, target tracking system, storage medium and electronic equipment Technical Field The present application relates to the field of image processing, and in particular, to a target tracking method, a target tracking system, a storage medium, and an electronic device. Background The intelligent road side parking management system generally performs in-out vehicle target matching by capturing behavior information of vehicles driving in and out of a parking space and vehicle identity information, forms a complete parking evidence taking order and realizes automatic road side parking charging. Because of its charging nature, multiple beats and missed beats of the target vehicle can affect the vehicle tracking management efficiency and accuracy. Particularly, when the vehicle is shielded by other vehicles or objects, the situation that the tracking target is lost easily occurs, the robustness of real-time tracking is reduced, and the tracking management of the vehicle is not facilitated. Disclosure of Invention The application aims to provide a target tracking method, a target tracking system, a storage medium and electronic equipment, which can effectively track an occluded target. In order to solve the technical problems, the application provides a target tracking method, which comprises the following specific technical scheme: Acquiring video stream data; detecting whether an occluded target exists in the video stream data; If yes, judging whether the blocked target can be effectively detected; if the blocked target is effectively detected, judging whether the upper boundary of a target detection frame of the blocked target is displaced or not; if the upper boundary is displaced, keeping track of the blocked target; And if the upper boundary is not displaced and the lower boundary is moved, performing a frame locking operation from the current frame to track the blocked target, wherein the frame locking operation is used for fixing the size and the position of a tracking frame. Optionally, detecting whether the occluded object exists in the video stream data includes: judging whether video frames exist in the video stream data, wherein the intersection ratio between a target detection frame and a suspected shielding object detection frame is larger than zero, and the vertical axis coordinate of the lower boundary of the target detection frame is smaller than that of the lower boundary of the suspected shielding object detection frame; If yes, confirming that the blocked target exists in the video stream data. Optionally, determining whether the upper boundary of the target detection frame of the blocked target is displaced includes: Determining a starting frame number of the blocked target in a static state; Determining the center point coordinates of the target detection frame in the video frame corresponding to the initial frame number; If the difference value between the center point coordinate and the second center point coordinate of the target detection frame in the next video frame is smaller than a set threshold value, determining that the upper boundary of the target detection frame of the blocked target is not displaced; And if the difference value between the center point coordinate and the second center point coordinate is not smaller than the set threshold value, determining that the upper boundary of the target detection frame of the blocked target is displaced. Optionally, after the determining whether the blocked target can be effectively detected, the method further includes: if the blocked target cannot be detected, judging whether the blocked target is contained in a low-resolution detection frame with the confidence coefficient lower than a preset value; if the blocked target is contained, continuing to track the blocked target; And if the low-resolution detection frame does not contain the blocked target, executing the frame locking operation from the current frame. Optionally, before judging whether the blocked target is included in the low-resolution detection frame with the confidence coefficient lower than the preset value, the method further includes: and performing target detection on the video stream data, taking a detection frame with the confidence coefficient larger than the preset value as a high-score detection frame, and taking a detection frame with the confidence coefficient lower than the preset value as a low-score detection frame. Optionally, after the video stream data is acquired, the method further includes: Recording the position information of the center point of the detection frame corresponding to the target; Fitting according to the position information of the central point to obtain a detection frame track; track prediction is carried out on the target by utilizing a tracking prediction algorithm to obtain a predicted frame track; determining an included angle between the detection frame track and the prediction frame track; and constructing