A Small Target Detection Algorithm Based on Immune Computation and Infrared Background Suppression

摘要

This research paper investigates a target extraction method of infrared point targets and proposes a type of space-time combined target detection algorithm. Starting from the detected target information which has a low signal-to-noise ratio and with very few target characteristics from a small target infrared image formation, it uses an algorithm for least squares involution filtering of the space with changing self-adapted restraints to process single frame images and suppress any background noise. In processing the images in this way a Gaussian white noise sequence is obtained. Then according to the relevance and continuity of the target movement and by adopting an intelligent path matching technology based on immunity computation the target is detected and the moving target path is obtained.