On particle filter and Mean Shift tracking algorithm based on multi-feature fusion

Abstract

To solve the problem that a single feature lead to tracking failure easily in a complex environment, an efficient particle filter and Mean Shift tracking algorithm based on multi-feature fusion was proposed. Under the framework of particle filter, it the closer to the real posterior distribution by embedding Mean Shift algorithm and using color and structural as the observation model to represent the object, and the weights of particles were calculated by this integration, in order to avoid the single color features easy to track the failure problem. The experiments show that the proposed method has a better robustness when using the same particles and the average weight of the particle is improved and the resample times reduced significantly, even using the less particles can achieve tracking stability.