Automatic Registration Method of SAR and Optical Image Based on Line Features and Spectral Graph Theory

Focusing on the automatic image registration problem of SAR and optical image because of the inconsistency of radiometric and geometric properties, a new algorithm based on line features and spectral graph matching is presented in this paper. Firstly, different edge detectors are employed to detect the line segments in both optical and SAR images respectively. With the random sampling consensus method, corresponding point pairs are found with these candidate points. Then voronoi diagrams are generated from two point sets. Finally, combined with spectral graph theory, the corresponding features can be obtained. Experimental results show the proposed method is more insensitive than traditional registration methods, as well as high precision performance.