A novel SVM classifier for recognition of gray character using Gabor filters

In this paper, we addressed the issues in feature of the gray character. Second we construct the designing a novel SVMs for gray character recognition. And in the stage of feature selection, we use the gabor filter which can scale the image source from different directions and different scales. In the stage of designing the classifier, we use the SVMs, however we construct each SVM according to our valuable features and consider the current samples just have two classes. And following our principle, we got the combining SVMs. All the process of recognition is like a binary tree.