Abstract: Improvement of high resolution image fusion by Mallat algorithm for orthogonal wavelet transform and stationary wavelet transform（SWT）was studied.With QUIKBIRD image as example, fusion test was made and compared with standard HIS and BROVEY methods.Comparison was made on the information volumes of the images before and after the fusion by using the indexes of information entropy and standard variation.Average value, correlation coefficient and standard bias index were used in the quantitative evaluation.Analysis was made on panchromatic band and multi-spectral band by using the correlation of detailed information with high resolution of fusion.The results indicate that wavelet transform fusion algorithm is on the whole better than the conventional ones and can well keep the spectral characteristics.

Key words: High resolution image, Stationary wavelet transform, Image fusion