A CF-based algorithm for moving dot target detection in FLIR image sequence

Dot target detection in FLIR image sequences with very low SNR has been used in guidance and surveillance systems and many algorithms have been developed to solve this problem. In this paper, a certain factor (CF) based algorithm was proposed, It is composed of two components: a predetector and a trace matcher. The predetector implements a N- P test on each pixel in the images. The trace matcher matches the detected dots in image with traces. Two certain factors were introduced to traces. Performance analysis was presented. It is shown that this algorithm has not only high detection probability but also low false alarm rate (if SNR=3, then PD=0.97, PF=10/sup -8/). Monte-Carlo experiments have bean conducted and the result is concordant with the theoretical analysis.<
>