Robust Infrared Small Target Detection Using Hough Line Suppression and Rank-Hierarchy in Complex Backgrounds

Xiaojie Liang,Lili Liu,Man Luo,Zujing Yan,Yunhong Xin

DOI: https://doi.org/10.1016/j.infrared.2021.103893

IF: 2.997

2021-01-01

Infrared Physics & Technology

Abstract:Recently, infrared patch-image (IPI) model has made breakthrough progresses in infrared small target detection. However, IPI model is degraded by inherent defects of the l1-norm and the inaccurate estimation of background patch-images, which causes sparse clutter edges around targets in backgrounds with heterogeneous structures. Herein, we introduce new strategies and techniques that improve the detection performance of the IPI model. We first propose a technique called Hough line suppression, which considers the difference between the features of strong edges and small targets, and maps the traditional global edge suppression problem onto a new local peakdetection problem in Hough space. This technique suppresses the background residuals while keeping the target largely intact. Next, we present a rank-hierarchy model that organizes the target candidates by the maximum differences between pairs of singular values. The rank-hierarchy further suppresses the residual clutter and accurately distinguishes the target from the pixel-sized noises with high brightness. Finally, we design a dualpath processing strategy that detects the absence or presence of Hough lines in an image, and selects the correct processing path. This strategy effectively improves the detection performance of images without Hough lines, and guarantees the detection performance of images with various backgrounds. Extensive experiments in various complex scenarios proved that the proposed method is more robust and effective than 12 state-of-the-art contrast methods.