Salp swarm algorithm based on golden section and adaptive and its application in target tracking

In order to solve the problem that conventional tracker is not adapted abrupt motion, a tracking algorithm based on improved salp swarm (ISSA) was proposed. Visual considered be process of locating optimal position through interaction between leaders and followers in successive images. Firstly, adaptive mechanism leader follower introduced into original (SSA) balance exploitation exploration algorithm. This method can improve accuracy effect tracking. Secondly, golden-sine used update followers, considering SSA had single spatial search mode for easy fall local optimum. By comparing with 19 classical algorithms, qualitative quantitative analysis carried out verify proposed method. A large number experimental results show here has good performance visual tracking, especially mutation motion