The fusion and evaluation of WorldView-2 images — A case study of general survey project of geographic national conditions in Sichuan, China

摘要

WorldView-2 image data is an important data source for general survey project of geographic national conditions in China. In this research, eight methods, namely Brovey, Ehlers, HPF, HIS, PCA, Wavelet, Pan-Sharpen and Gram-Schimdt were adopted to fuse panchromatic and multi-spectral images of WorldView-2. Qualitative analysis from the aspects of clarity, texture and tone, and quantitative evaluation from the aspects of standard deviation, information entropy, average gradient, deviation index, correlation coefficients, and spectral distortion degree had been done. The results show that the image distortion degree based on the transformation method of Pan-Sharpen is smaller, and meanwhile, the fused image can largely keep the spatial texture details of high-resolution panchromatic band, so it is a suitable method for fusion of WorldView-2 images.