A Sequential Frame Method for Traffic Sign Detection

Based on the sequential video frame, a method to detect the traffic sign was proposed. The color image was mapped from RGB to improved HSI color space first, and then converted it to binary image according the color of signs. The shape features were used to remove some noise, and the improved Hough transform was used to locate the areas of traffic signs. Compared to the existed detection algorithm, a previous frame was used as a priori information to reduce the treatment region, then to decrease the computation. In order to prevent the undetected region, a full inspection was processed when the number of consecutive frames which detected the traffic signs was more than the threshold. The results show that the method in this paper performs better than the traditional method.