Hyperspectral image classification using kernel method based on the correlation coefficients of neighbor bands

Based on the framework of support vector machines (SVM) using one against one (OAO) strategy, a new kernel method based on the correlation coefficients of neighbor bands is proposed to raise the classification accuracy by combining the characteristics of hyperspectral image. This algorithm assigns weights to different bands in the kernel function according to the amount of useful information that they contain, which makes the band with more useful information play more important role in the classification. Our research has shown that the band with greater the correlation coefficients between neighbor bands contains more useful information, and hence we use the correlation coefficient of each band and its neighbor bands as the weights of the proposed kernel method. The experimental results show that the support vector machines based on the correlation coefficients of neighbor bands is effective and feasible, and the numbers of the support vector reduced to some extent.