Weak point target detection in the complicated infrared background

摘要

Point target detection is an important issue for the application of infrared search and track (IRST) system. To detect the weak point target in infrared image, this paper introduces a new method based on two-steps gradient algorithm to eliminate the impact of complicated background and extract possible targets. Firstly, the characteristics of point target are analyzed based on 288×4 linear infrared FPAs. Secondly, the point targets in complicated infrared background are classified into four groups. The detection function is derived for each group, and then the comprehensive detection function is derived to extract the point target. Finally, a practical experiment is completed based on a 288×4 IRST system. It is proved that the two-steps gradient algorithm can reduce the number of possible targets which is helpful to the post-processing work.