Abstract: Objectives:To address the challenging problem of matching heterogeneous remote sensing images caused by nonlinear radiometric distortions,this paper proposes a nonlinear scale-space enhanced automatic matching method for optical and SAR images. Methods: First,by improving the calculation of color pixel contrast,the contrast information of the images is effectively enhanced,improving the contrast consistency between optical and SAR images,and enhancing the repeatability of corresponding points. Then,a nonlinear diffusion equation is employed to describe the image diffusion characteristics,avoiding the issue of boundary blurring in the Gaussian scale-space. Subsequently,the Multiscale Ratio Of Exponentially Weighted Averages (ROEWA) operator and the Sobel operator are utilized to compute the gradient information of the SAR image and the optical image,respectively,followed by the extraction of stable Harris feature points. Finally,a joint Log-polar descriptor framework is employed to compute a high discrimination feature vector,and outliers are eliminated using the Euclidean distance and the Fast Sample Consensus (FSC) algorithm.Results and Conclusions: Experimental results demonstrate that compared to the PSO-SIFT,OS-SIFT,and HAPCG methods,the proposed algorithm achieves a higher number of matched corresponding points while maintaining similar accuracy,thereby achieving automatic and robust matching of SAR and optical images.