The Segmentation and detection of will target image for infrared homing guidance head are crucial technolyies for ground-air and surface-air missiles. This paper studies the effect and performance coparison of suppnaing largearea and low frequency background by use of six high-pass filters. Based on this

we propose an adaptive threshold method ofbackground supprasion and small target segmentation using image mean and histogram. In view of the continuity and regularity of motion for target and the randomnere for noise

the design of continuity filter and a technique via assembly line andpipeline scheme to detect target trajectory are considered. The methods rourched here are verified by a large number of resultS of experiments and comparisons.