Remote-Sensing Image Fusion Based On K-means Algorithm And Robust Regression

Image fusion means to make the best use of multiple source pictures and get more accurate and more comprehensive information about the same object or scene. Multispectral images have abundant spectral information, well panchromatic images have higher spatial resolution, along with the development of remote sensing technology, how to infuse the multispectral and panchromatic image in order to get a image with higher spatial resolution and abundant spectral information is a key point. To achieve this goal, in this paper, it introduces the K-means algorithm and the robust regression.In this paper, it introduces the basic information of image fusion, the concept, purpose of image fusion are also involved. Image fusion contains three levels. It focuses on the methods and evaluation of pixel fusion method, such as the principle component analysis (PCA) and the wavelet transform. Some qualitative and quantitative criterions are also introduced in detail.In this paper, it combines the K-Means cluster with the robust regression together, not only guarantees the linear relation between the multispectral and panchromatic image, but also reduces the outliers. A series of experiments on image fusion are given in this dissertation. The experimental and calculation results show that the fusion algorithm we used is effectual.