DP–MHT–TBD: A Dynamic Programming and Multiple Hypothesis Testing-Based Infrared Dim Point Target Detection Algorithm

The detection and tracking of small targets under low signal-to-clutter ratio (SCR) has been a challenging task for infrared search track (IRST) systems. Track-before-detect (TBD) is widely-known algorithm which can solve this problem. However, huge computation costs storage requirements limit its application. To address these issues, dynamic programming (DP) multiple hypothesis testing (MHT)-based dim point target (DP–MHT–TBD) proposed. It consists three parts. (1) For each pixel in current frame, the second power optimal merit function-based DP designed performed eight areas to find area that contains real trajectory. (2) In area, parallel MHT model save tree-structured trajectory space, two-stage strategy mitigate contradiction between redundant trajectories more SCR. After constant false alarm segmentation energy accumulation map, preliminary candidate points be obtained. (3) method eliminate alarms. work, an efficient pixel, greatly reduces space. A model, pruning space avoided all processed parallel, further reduce exponentially. This computational complexity saves improving engineering application TBD method. DP–MHT–TBD not only takes advantage amount high accuracy exhaustive but also utilizes novel structure. detect single when SCR 1.5 with probability above 90% rate below 0.01%.