Kalman Filter Based Point Target Tracking

摘要

The detection and tracking of point targets in cluttered environment has always played a significant role in surveillance and navigation applications especially for aerial targets. Most of the contemporary MWS (Missile Warning Systems) and IRST (Infra Red Search and Track) systems use captured IR radiations from the scene for detection and tracking of such point targets. The unintended IR (Infra Red) radiations emitted from targets make the signatures more prominent than background and clutter, this contrast is commonly used as a clue for detection of targets. The present approach also relies on this contrast based clue. On the basis of statistical analysis of image content of row and column vectors and Kalman filtering, the location of the point target is explored in the considered frame of the video sequence. A dataset of synthetic video sequences having point target with variable SNR is generated to validate the proposed method. The results obtained demonstrate effectiveness and robustness of the approach.