Aiming at color video sequences

conventional kernel correlation filtering（ KCF） tracking method does not take the color information into consideration and lack of occlusion processing scheme

when the target is under the influence of occlusion

illumination changes

etc.

it can not handle these very well. In order to solve this problem

combined with the HSV spatial color information

a tracking method using multi-feature fusion and the losing handling is proposed

and adopt the method of re-detection to deal with the problem of losing of the target.23 challenging video sequences in the public test video sets are selected

after the test

compared with the classical KCF method

the mean center location error（ CLE） is reduced by 29. 51 pixels and the average distance precision is improved by 23. 87 %. The experimental results show that the proposed method has strong robustness while the target is under complex conditions such as fast moving

shape changes

occlusion and rotation.