Traffic Sign Recognition Using Scale Invariant Feature Transform and Svm

TSR (Traffic sign recognition) has been studied for realizing drivers assisting system and automated navigation and is an important studied field in ITS (Intelligent traffic system). In this paper, a recognition method of traffic signs separated from real image was studied. Images were divided into several categories according to the actual weather, distance and angle of view etc. SIFT was firstly used to detect keypoints and describe them because the SIFT(Scale Invariant Feature Transform) features were invariant to image scale and rotation and were robust to changes in the viewpoint and illumination. And then the Bag-of-words method was applied to recognition. Finally, the SVM(Support vector machine) approach was used in classification. As shown by the experiment results, the proposed method is effective. * Xiaoguang HU , Ph.D candidate, majors in pattern recognition,GIS and computer vision etc. [email protected].