干旱区生态环境知识资源中心

One of the crucial steps in the fusion of remote sensing image is the selection of proper fusion techniques. In this paper, we focused on the bad geology, such as desert, permafrost, saltmarsh, etc., in Yuli Rob Village, Xinjiang Province. The selection of best band in remote sensing images is treated as a combinational optimization problem, and an optimum index factor method is proposed for the selection of best band in remote sensing images. Based on the analysis of general fusion techniques used in remote sensing image, three representative techniques, i.e., Gram-Schmidt, color normalized transformation, and principle component transformation are used for the fusion of bad geological remote sensing images. Experimental results indicated that the techniques used in this study improve the image spatial information, and they can maintain the spectral feature of multi-spectral images.