Abstract:A new approach for detecting moving dim targets in infrared image sequences is proposed. The approach adopts an innovative density-based cluster to eliminate the interferences of heavy and slowly evolving clutters and extract pulse signals on temporal profiles that are generated by moving targets. Two infrared senses are processed by this approach and other conventional approaches. The results show the approach has high robustness to heavy and slowly evolving clutters.

Moving Dim Target Detection Via Density-Based Cluster in Infrared Image Sequences