An experimental consideration for road guide sign understanding in ITS

In this paper, we present a road guide sign recognition method the driver support system in intelligent transportation systems (ITS). Generally, on a guide sign, there are arrows representing directions and characters representing destinations and distances. First, arrows are detected using their shape characteristics. Next, the sign is classified using the number of arrow objects and the number of arrowheads. Then, the characters on the sign are assigned to their corresponding arrowheads by segmenting the sign using arrowhead positions. Finally, after the character recognition process, we obtain the information of directions, destinations and distances. We tested 65 images of zoomed in signs. 154 arrowheads on these signs were detected completely. In 62 images, characters were assigned to corresponding arrowheads successfully. Hence, our method is effective for sign image understanding.