Missile vision guidance based-on adaptive image filtering

A novel target tracking algorithm based on adaptive image filtering is proposed to realize missile vision guidance. According to the specific condition of application, we propose an adaptive filtering, light and effective motion model, and a fusion algorithm. Small target scale in the first, large scale variation during the procession, and the changing complex backgrounds can all be solved effectively by our algorithm. Simulated experiments show that our algorithm performs well with 15% distance error ratio, and more than 70% overlapping in average. The speed of real-time procession is faster than 300FPS.