Coarse to Fine Level Set Segmentation of SAR Imagery Based on The Brovey Transform Fusion of Optical Imagery

摘要

This study delivers a new approach for contour extraction in different terrains of the same location of both SAR images and optical images using a coarse to fine level set scheme. Segmentation in both SAR and optical imagery is a trivial challenge for many researchers. The main target is to segment buildings, lands and oil tanks perfectly via fusion with a level set method. First Brovey transform is applied in fusing both SAR images and optical images. Discrete wavelet transform is utilized for extracting an image from a background. Finally in order to segment an image a coarse to fine level set is implemented. Experiments have been carried out on quick bird and Terra SAR-x images to validate the proposed methodology.