Intuitionistic fuzzy random forest based target tracking method and apparatus

Abstract

The present invention discloses an intuitionistic fuzzy random forest based target tracking method. The method comprises: carrying out movement detection on a current video frame, and taking possible movement objects obtained from detection as observation results; associating the observation results with prediction results of targets, wherein the targets comprise reliable targets and temporary targets; carrying out trajectory management on unassociated observation results and prediction results, including carrying out on-line tracking on the unassociated prediction results of reliable targets so as to obtain candidate results, and using intuitive fuzzy random forests of the unassociated reliable targets to match the candidate results; and using associated results and matched results to obtain a trajectory of the target of the current frame, using the trajectory of the target of the current frame to predict, and updating the intuitionistic fuzzy random forest for the successfully associated or successfully matched reliable targets. The present invention further discloses an intuitionistic fuzzy random forest based target tracking apparatus. According to the method and apparatus disclosed by the present invention, performance of the target tracking in the case of the leak detection can be improved.

Description

technical field [0001] The invention relates to the field of target tracking, in particular to a target tracking method and device based on intuitionistic fuzzy random forest. Background technique [0002] Online object tracking is a hot research topic in computer vision. It is of great significance to high-level visual research such as action recognition, behavior analysis, and scene understanding, and has a wide range of applications in video surveillance, intelligent robots, and human-computer interaction. prospect. [0003] In complex scenes, due to the influence of factors such as the deformation of the target itself, the mutual occlusion between the targets, or the occlusion of the target by the background still life, it is inevitable that missed detection will occur. At this time, the missed detection objects cannot find the detected observation objects associated with them, and it is impossible to find effective information for the trajectory update of these missed ...

AI Technical Summary

Problems solved by technology

[0004] The technical problem mainly solved by the present invention is to provide a target tracking method and device based on intuitionistic fuzzy random forest, which can solve the problem of reduced trajectory accuracy of missed targets in the prior art

Images