Infrared target tracking based on repair particle swarm optimization

To avoid the particle swarm algorithm degradation and complicated computation, a repair particle swarm algorithm for infrared target tracking is introduced. Firstly, the algorithm uses the inertia factor of the defined particles to correct the location of the infrared target detected, so that the location of particles reaches local optimum and global optimum. Secondly, PSO constriction factor is limited to border search to eliminate the ambiguity of the target location. Results show that there is an error of 2.83% in 500 granule iterations and 100 tracks. The tracking result is closest to real target with sharp edge when the maximum inertia weight is 1.2 and the minimum inertia weight is 0.3.