Fusion of Multi-Sensor SAR Images via Adaptive Selection of Wavelet and Contourlet Coefficients

Using SAR images of different spectral bands relative to the same scene can obtain a better knowledge of the scene than the one obtained using only one image. Fusing these images can provide more comprehensive information. In recent years, the fast growth of multiscale geometric analysis (MGA) has brought out abundant tools for image processing, such as curvelet and contourlet transform. In this paper, a novel fusion method based on adaptive selection of wavelet and contourlet coefficients is proposed for SAR image fusion. The main merit of the proposed method is that the wavelet transform can locate sparse of spot strangeness of SAR image effectively and contourlet transform can give a sparse representation of line strangeness, thus it can combine advantages of these tools, and get more details than other methods. The experimental results of four SAR images demonstrate that this algorithm performs better than classical fusion methods both in visual effect and object evaluation