Multi-source Remote-sensing Image Fusion Based On Wavelet Analysis And Unsupervised Classification

Data fusion of remote-sensing images is one of the most interesting problems in the field of remote sensing technology. With the increase of the means to acquire remote information and the development of remote-sensing technology, remote-sensing image fusion has attracted more and more attention in recent years. In this paper, theories about wavelet analysis and image fusion are reviewed and remote-sensing image fusion is realized using wavelet analysis and unsupervised classification.First, wavelet transform theories are systematically reviewed and summarized from the point of view of signal and information processing. 2-D orthogonal wavelet transform and redundant wavelet transform are realized by use of Mallat algorithm and Trous algorithm respectively. Then typical existent fusion algorithms based on orthogonal and redundant wavelet transform are summarized. Standard deviation, entropy and average grads are advised to be the objective evaluation parameters. And the characteristics of each algorithm are analyzed theoretically. According to this analysis, a new pixel-level fusion algorithm based on the activity and similarity of source images is proposed and its performance is tested to be superior to the congeneric algorithms. By in-depth research of image texture and its application in multispectral image fusion, Significant Central Coefficient (SCC) algorithm based on redundant wavelet texture is proposed and its performance is tested to be also superior to the congeneric algorithms in the way of enhancing fusion quality. In the end, multispectral image fusion algorithms based on Principal Component Analysis and clustering algorithms are introduced. Multispectral image fusion based on unsupervised classification is realized.