Traffic Sign Classification Using Ring Partitioned Method

--Traffic sign recognition usually consists of two parts : detection and classification. In this paper we describe the classification stage using ring partitioned method. In this method, first the RGB image is converted into gray scale image using color thresholding and histogram specification technique. This gray scale image, called as specified gray scale image is invariant to the illumination changes. Then the image is classified using ring partitioned method. The image is divided by several concentric areas like rings. In every ring the histogram is used as an image descriptor. The matching process is done by computing the histogram distances for all rings of the images by introducing the weights for every ring. The method doesn’t need a lot of samples of sign images for training process, alternatively only the standard sign images are used as the reference images. The experimental results show the effectiveness of the method in the matching of occluded, rotated, and illumination problems of traffic sign images.