A Method of Small Moving Target Detection

摘要

This paper present a new method for detecting small moving targets from sequential images under the complex background of sky. The method can identify moving targets in low SNR, estimate and track the trajectory of the target using the maximum gray matrix and Kalman filter theory. Compared with other algorithms, this method has features of small amount of calculation, fast detecting speed, high real-time performance, and small identification error. Experimental results show that the pepper noise and fixed pattern noise in this type of image sequence have a good filtering effect, and also our method has good recognition and tracking effect for the small moving targets.