Efficient small infrared (IR) maritime target detection in heavy waves is a key and challenging task in maritime distress target search and rescue systems. The current methods are struggling to enhance targets and suppress heavy waves because of the similar local contrast. In this letter, a weighted multidirectional gradient (WMDG) measure is presented. First, the candidate targets are extracted via facet model. Then multidirectional gradient and difference information are constructed to calculate average cumulative multidirectional gradient (ACMG) and achieve directional difference measure (DDM). Consequently, the final saliency is reconstructed via ACMG weighted by DDM. Finally, the actual targets are segmented through an adaptive threshold. Experimental results demonstrate that our method is superior to the state-of-the-art methods with regard to detection validity and robustness for small IR maritime target detection in heavy waves.
Read full abstract