This article deals with automatic classification methods and applying of digital image. In this paper, put forward a method that using MGabor filter-banks to extract Texture-Color features from images, and then constructing a group of support vector machines (SVM) classifiers to automatically and accurately classifying color digital images. Successful experiments to Simplicity and Brodatz Image set and our own ancient shards image sets. The experiments results show our methods that integration of the texture features and color information can further improve distinguishing ability of each category images.
展开▼