Global self-localization of redundant robots based on visual tracking

In order to improve the global self-localization effect of redundant robots, based on the visual tracking algorithm, this paper studies the global positioning method of redundant robots and constructs corresponding models to improve the working efficiency of redundant robots. Moreover, this paper uses the CamShift tracking algorithm combined with the Kamlan predictor to track and predict the dynamic image, and introduces the epipolar constraint in binocular stereo vision to constrain the tracking particles, which improves the matching accuracy of the two tracking particles. The global self-positioning system software is mainly composed of two parts: the robot chassis software and the global self-positioning ontology software. Finally, this paper combines the actual situation to construct a redundant robot global self-positioning system structure based on visual tracking, which can provide a reference for subsequent related research.