Research on Integrated Monitoring Algorithm for Small Water Bodies in Complex Terrain Regions Using Sar and Optical Fusion

Abstract

The spatiotemporal distribution of small water bodies is crucial for water resource management and policy formulation. However, existing studies often rely on a single remote sensing data source (optical or microwave), focusing on large-scale, flat areas and lacking comprehensive monitoring of complex terrains and scattered small water bodies. Therefore, considering the complementary advantages of multisource remote sensing (multispectral and SAR), this paper proposes a multispectral and SAR fusion algorithm (MASF) based on a multi-resolution analysis framework. The MASF first utilizes weighted least squares filtering to decompose optical images, optical-guided filtering to decompose SAR images. Subsequently, it calculates the weights for spectral and spatial components. Finally, the fused image is obtained after spectral compensation and inverse transformation, which enhancing spatial features and improving spectral fidelity. Based on this, a dataset containing spectral, texture, and geometric features is generated using the fused images. Further, the study combines multiscale segmentation with a random forest algorithm to achieve precise identification and extraction of small water bodies (500 square meters, equivalent to 5 pixels) in complex terrains. The results demonstrate that the proposed fusion algorithm, MASF, exhibits minimal spectral distortion (SAM < 3.5, ERGAS < 21, RMSE < 0.01, CC > 0.99) and robust spatial feature enhancement (PSNR > 40, SSIM > 0.999). The overall accuracy and Kappa coefficients for rivers and reservoirs in the experimental area are both above 90%. However, the overall accuracy of agricultural ponds is below 80%, primarily due to the fragmented distribution patterns of agricultural ponds in complex mountainous terrain. Our study addresses the limitation of relying solely on spectral information to identify different small water bodies, thereby providing crucial support for water body conservation in regions with cloudy and complex terrain conditions.

Keywords: SAR and optical image fusion, small water body, complex terrain regions, multiscale segmentation, Random Forest