Image fusion algorithm of the hyper-spectral remote sensing appling to the survey of land utilization based on a curvature-wave transformation
曲波变换的高光谱遥感图像融合方法在土地利用调查中的应用

In this paper, after several-hundred times ofcomparing calculation of the hyper-spectralandQuickbird data in two land utilization survey areas, The Sym4andDb2are determined to be the bestwavelet function in the curvature-wave transformation at las.t Two hyper-spectral remote-sensing image fusion algorithms, .i e., weightedmethod and selectedmethod, are given based on the curvature-wave transformation. And then themore precise fusion images are provided. In one research area, the hyper-spectral remote-sensing fusion image is benefited for the land typing and image discrimination after the images to resolve son, ridge-wave transformation, some fusions, and their inversion transformation of the hyper-spectraldata. In the other research area, exceptfor the curvature-wave transformation, the remote-sensing image matching of the hyper-spectral and the Quickbird data is finished by calculation of quadratic equation, and the fusion image of the curvature-wave transformation is better than that of wavelet transformation, and similar to theBrovey fusion image and slightly inferiorofPCA fusion image after the appraising by observation, information entropy and relation analysis. It is suggested that the hyper-spectral remote-sensing image fusion algorithm can provide more precise information for the land utilization survey based on the curvature-wave transformation.