Infrared Small Target Detection Based on Adaptive Balance Structure Tensor Indicator

Infrared small target detection has extensive applications in various systems, including search and tracking, guidance and missile defense, as well as electro-optical pods. Although the existing saliency-based detection methods have demonstrated promising performance, two key issues include the inability to effectively suppress continuous strong edges and the lack of robustness in manually adjusting thresholds, persist as ongoing challenges. To address the aforementioned issues, we propose an infrared small target detection method based on a structure tensor indicator, which incorporates a structure tensor-based edge indicator with an adaptive balancing weight. Initially, structure tensor features are utilized to design two indicators that separately characterize clutter edges and targets, while incorporating a balancing weight to enhance the target and mitigate edge interference. To overcome the lack of adaptability in manually configured parameters, we propose an adaptive weighting factor based on a structural similarity index measure, which iteratively adjusts the weight by evaluating the differences between the target image and the detection image. Extensive experiments conducted on real-world datasets underscore the superiority of the proposed method in terms of target enhancement, background suppression (BS), and detection efficiency.