Target tracking method with self-restoration capacity based on multi-stage detector

Abstract

A target tracking method with self-restoration capacity based on a multi-stage detector includes the steps of selecting a plurality of detectors of different types to be connected in series in combination with the concept of cascading Adaboost multi-stage weak classifiers into a strong classifier, selecting a significance detector for first-stage detection, selecting a classifier assembly detecting module for second-stage detection, substituting the classifier assembly detecting module into a random tree to calculate the posterior probability of a positive sample in a probable area detected byfirst-stage detection, selecting a related filtering detector for third-stage detection, calculating the relevance between a sample with the posterior probability larger than a certain threshold andthe positive sample initialized or obtained in the last frame so as to reduce accumulated errors caused by long-term tracking, determining the position with the maximum relevance as the target area inthe current frame through the multi-stage detector, sampling the determined position, supplementing the positive and negative sample number of concentrated removed samples to ensure the reliability and number consistency of the samples, activating a redetection mechanism if the maximum value of the relevance is larger than a certain threshold, and detecting the area near the position again to search for a target.

Description

technical field [0001] The present invention designs a target tracking method with self-recovery capability based on multi-level detectors, and connects multiple detectors of different types in series to form a strong detector to detect the target on each frame of the video sequence, thereby realizing the target Tracking, and can perform re-detection when the tracking target is lost, re-detect the target position and continue tracking. This method adopts the idea of ​​target tracking based on detection, which has good robustness to static background and dynamic background, and has high engineering application value. Background technique [0002] Theoretical research on how to maintain long-term stable tracking of image targets has gradually become a direction of higher concern in computer vision, which has been widely used in monitoring and recognition and robotics and other fields. Object tracking is a comprehensive research and development direction that integrates many d...

AI Technical Summary

Problems solved by technology

However, in the actual process of using detection to track the target, the common problem of the above algorithms is the tracking error caused by various possible factors. During the tracking process, the error gradually accumulates, which will lead to the decline in the accuracy of the subsequent detection target, and finally Affect Tracking Effect

In addition, in the initialization stage of target tracking, the acquired target initial template itself has accuracy problems, which will seriously affect the accuracy of the subsequent tracking process.

If there are obstacles or buildings in the trajectory of the target's movement, the target's features will disappear partially or even completely due to being blocked. When the target reappears, it will not be able to re-detect and continue tracking

Images