Regional Guidance-based Multi-Scale Joint Filtering Model for SAR Images

Synthetic Aperture Radar (SAR) imaging systems are characterized by their all-weather and day-and-night capabilities, enabling the acquisition of high-resolution ground imagery. However, the presence of speckle significantly hinders the effective utilization of SAR images. In this paper, a novel coherent speckle suppression method was proposed, leveraging the statistical characteristics and texture information inherent in SAR data. To account for the texture features of SAR images, a new guide map—termed Regional Guided Images (RGI)—was developed and integrated into the Non-Local Means (NLM) framework to implement a region-based speckle reduction approach. Double hesitant neutrosophic set functions were constructed by incorporating both the statistical distribution and multi-scale features of SAR images. In addition, surface variation was introduced to model the non-local self-similarity properties of SAR data. By embedding this framework into non-local filtering models, the proposed method effectively achieved coherent speckle suppression while preserving structural details. Comparative experiments on real SAR images demonstrate that the proposed method outperforms several state-of-the-art techniques. It achieved a favorable balance between speckle reduction and edge preservation, as evidenced by both visual assessments and quantitative evaluations. These results verify the effectiveness and advancement of the method in SAR image speckle suppression.