Infrared Small Target Detection Based on Patch Image Model with Local and Global Analysis

Patch image model has recently shown significant superiority in the detection of infrared small and dim targets. In this paper, we incorporate more useful local and global information into the sophisticated patch-image model called reweighted infrared patch-tensor model, for its efficiency and flexibility. Local signal-clutter-ratio analysis is employed to enhance targets and avoid targets being overwhelmed by strong background edges. In the meantime, nuclear norm minimization is applied to globally measure the low-rank property of a couple of background matrixes generated from all the patch-mages. Also, noise patch-mages are identified by adding an l(21) norm in order to deal with the rare structure effect. Experimental results show that the proposed approach endows high detection probability and robustness to noise, and outperforms state-of-the-art methods in complex scenes.