Application of a Feature Point Tracking Method for an Infra-red Seeker

摘要

We present in this paper an algorithm for target tracking applying a feature point tracker, Kanade-Lucas-Tomasi tracker, to the Imaging Infra-red seeker. This research focus mainly on generating the new point starting to track in the next frame in case of losing the current tracking point during. Assuming that at least one of some reference points around the current tracking point has the same movement with it, we select one of them as a tracking point in the next frame. We also show the result of correlation based tracker and using kalman filter to compare the performance .in experiments.