Multi-feature and self-adaptive standard hedging combined target tracking method

Abstract

A multi-feature and self-adaptive standard hedging combined target tracking method is characterized by comprising the following steps that (1) using multiple features of a tracking target to constructa dictionary, wherein the tracking target is expressed as a vector in a dictionary space; (2) designing self-adaptive standard hedging based on the standard hedging; and (3) in a self-adaptive standard hedging tracking framework, estimating the current tracking target. The multi-feature and self-adaptive standard hedging combined target tracking method still has good accuracy and robustness in acomplex environment.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a target tracking method combining multi-features and self-adaptive standard hedging. Background technique [0002] Visual target tracking is an important research content in the field of computer vision. It has been widely used in visual navigation, human-computer interaction, intelligent transportation, video surveillance and other fields. It is a variety of follow-up advanced processing, such as target recognition, behavior analysis, and application understanding. The basis for high-level video processing and applications. However, due to factors such as occlusion, illumination changes, scale changes, sudden changes, and angle changes in tracking videos, it makes accurate and robust video object tracking a very important task. [0003] With the development of compressed sensing theory and sparse coding theory, sparse representation has been applied to the tracking of ...

AI Technical Summary

Problems solved by technology

In sparse representation tracking, l 1 The tracking method has strong robustness, but because of its need to solve l 1 The norm minimization problem makes solving difficult and time-consuming

In view of this, Mei et al. adopted the minimum error limit to reduce the particles to increase the running speed, and Bao et al. adopted the accelerated gradient method to solve l 1 Norm, Zhang et al. proposed real-time compressed tracking methods to reduce the data dimension and increase the calculation speed, but when the illumination changes drastically or the target is similar to the background, the robustness and accuracy of these methods are significantly reduced

At the same time, there is an unavoidable particle degradation problem in the particle filter tracking framework, which is mainly manifested in that after several recursions, except for a small number of particles, most of the particle weights are small or even negligible, which will lead to tracking failure

Images