Traffic Sign Detection based on Color and Boundary Shape Box Ratio

This paper proposes traffic sign detection based on color and boundary shape box ratio. Red and blue traffic sign are separately extracted by color extraction which is done in HSV color model. Both red and blue candidate images are grouped and bounded in a shape based extraction process. The boundary ratio is applied to the red and blue candidate traffic sign image in order to extract the traffic signs from the image. The width and height ratio filtering is applied in order to eliminate the unwanted data in the image. Finally, the traffic signs are detected by combining red and blue candidate images. The experimental results show that the proposed method can detect 87 out of 99 traffic signs from 60 captured images.