Infrared small target detection serves as a cornerstone of infrared search and track systems, yet existing methods struggle with poor real-time performance and low detection accuracy under complex backgrounds. Accordingly, the paper proposes a fast infrared small target detection method based on weighted double dilation strengthened mechanism (WDDSM). First, inspired by the morphological characteristics of infrared small targets, the double dilation strengthened mechanism (DDSM) is established to amplify the saliency of targets, which leverages dilation operations. Second, to improve detection speed, a method for calculating the weighted coefficient (W) based on the candidate target points screened from the DDSM result is proposed, which effectively improves the detection efficiency while further suppressing background clutter. Finally, the DDSM result is fused with the weighted coefficient to generate a response map, followed by adaptive threshold segmentation to achieve accurate target localization. Experiments on 5 real infrared datasets against 8 baseline methods validate the superiority of our WDDSM. It achieves the highest BSF in 3 of the 5 datasets (peak 104.1544), optimal CG in 4 of the 5 datasets (max 1.8908), top AUC across all datasets, and fastest processing speed (as low as 0.01 s per frame) while well balancing detection accuracy and real-time performance.
Cui et al. (Fri,) studied this question.