Research on Identification Methods for Small Infrared Targets in the Air

This paper proposes a method for recognizing small infrared targets in the air based on Gabor features and sparse representation theory. Infrared weak and small targets are defined as targets exhibiting both “weakness” and “smallness” in imaging systems, characterized by low contrast, low signal-to-noise ratio, and a small number of pixels. Existing research primarily focuses on infrared small target detection, with limited studies on recognition technology. This paper fills this gap by exploring the feasibility of infrared small target recognition and proposing a recognition method. By simulating four types of aerial small targets smaller than 30x30 pixels, the method extracts Gabor features and constructs an over-complete dictionary using sparse representation theory. Experimental results show that the proposed method achieves an accuracy of 80.12 % in recognizing infrared weak and small targets, outperforming existing methods.