Target Detection and Tracking in Forward-looking Infrared Image Sequences Using Multiscale Morphological Filters

摘要

In this paper, we propose a robust approach for detection and tracking targets in forward looking infrared (FLIR) imagery taken from an airborne moving platform. Based on target intensity and shape information, motion criteria, the proposed algorithm involves three modules. Firstly, we use multiscale morphological filters to remove noise and clutter in tracking window. Secondly, target is detected by segment the image to produce a binary image. In the last step, our algorithm uses spatial-temporal connectivity to detect and track true target. The experiments performed on FLIR image sequences, show the robustness of the proposed approach.