Video object tracking method fused with tld and kcf

Abstract

The invention discloses a video target tracking method combining TLD and KCF. KCF algorithm modules run in parallel, if only one module of the processing TLD algorithm module and KCF algorithm module of the current frame has a tracking target output, then take this output as the tracking result of the current frame; if both tracking modules have tracking target output, then Calculate the similarity S between the tracking target output and the target model M respectively t and S k , choose S t and S k The largest among them is taken as the target tracking result; the next frame of video is processed through the above method until the end of video frame tracking. The method can overcome the shortcomings of the TLD algorithm and the KCF algorithm when used alone, has wide adaptability to the target tracking of complex video scenes, and maintains the real-time performance of the target tracking algorithm.

Description

technical field [0001] The invention relates to the technical field of video image processing methods, in particular to a video target tracking method combining TLD and KCF. Background technique [0002] Target tracking is the process of continuously obtaining the motion state of the target in the video sequence when the prior information of the target is known. It is one of the challenging research difficulties in the field of computer vision and has been widely used in various defense and civilian applications Fields, such as intelligent monitoring, human-computer interaction, medical diagnosis, navigation guidance, etc. Although a lot of research results have been achieved in video tracking technology in recent years, robust visual tracking is still a great challenge. The difficulties of video tracking mainly lie in the difficulty of accurately expressing target templates, complex target motion patterns, target deformation and scale changes, target occlusion and backgrou...

AI Technical Summary

Problems solved by technology

[0005] Although the TLD algorithm and the KCF algorithm have achieved good results in the tests of multiple video data sets, there are still obvious shortcomings when they are applied to the target tracking of actual complex scenes.

For example, the stability of the tracker of the TLD algorithm is not high, and it is easy to fail to track when there is complex background interference

The KCF algorithm cannot handle the situation where the target is occluded and the target is deformed.

Images