164 A New Speed Sign Recognition Algorithm Based on Statistical Characteristics

Traffic sign recognition which allows us to recognize signs in real time in videos is very important for road safety and driving behavior study. A new speed sign recognition method based on statistical characteristics is proposed in the paper. First, the selected region from road scene image which needs to be processed is determined and filtered by the SVF (Simple Vector Filter). Second, the speed signs are segmented after all the detected non-traffic signs pixels are excluded by the statistical characteristics. Then we are able to apply normalized cross-correlation to classify the signs. Potential signs are compared with the template signs as given in the database by using feature matching methods. At the end, we recognize the traffic sign in an image aiming at real time system. This method is suitable for circular signs only, we apply it to Chinese speed signs in this paper. Our results show that it is robust to a broad range of lighting conditions. The result of this method is about 96% for detection rate, and the speed sign recognition rate is ab out 90%.