SAR and multi-spectral images fusion for shallow-buried fracture detection in urban area

Abstract

Exploring the regional fractures in urban area and its surroundings is an important issue. Satellite remote sensing image can be used to illustrate the panorama of regional fractures and structures on macroscopic view. Fusion of multi-source images(especially SAR and optical multi-spectral image) is a hot spot in RS information processing and application field. A image fusion based linear structure information detection method is proposed in this paper. After introducing wavelet and wavelet packet transform and their applications in image fusion, a regional standard deviation wavelet coefficients fusion method is proposed. Experimental result shows wavelet packet fusion is better than wavelet fusion in resolution enhancement for fusion of Radarsat SAR and Landsat MSS, ETM+ multispectral remote sensing images. Study of exploring shallow-buried regional fracture information covered by Quaternary unconsolidated sediments in Hangzhou area shows the fusion method applied on SAR and optical multi-spectral RS image can improve resolution and the capability of interpret shallow-buried fractures.