Hyperspectral Image Classification Based on Empirical Mode Decomposition and Local Binary Pattern

Traditional hyperspectral image classification methods always focused on spectral information, and lots of spatial information was neglected. Therefore, this paper introduces the spatial texture information in the process of hyperspectral image classification, and focuses on how to deeply combine the texture information and the spectral information. Based on empirical mode decomposition and local binary pattern, the method of support vector machine is used to classify hyperspectral image, in order to improve the image classification accuracy.