Ground traffic signs recognition based on Zernike moments and SVM

In this paper, we propose an approach for recognizing ground traffic signs based on Zernike moments and support vector machine. Traditional traffic signs recognition was focus on the signs beside road, but there are relatively few researches on the ground traffic signs recognition. Ground traffic signs are important indication signs to indicate the carriageway. During the process of vehicle running, drivers need to judge which carriageway they are in, and select the right carriageway according to the itinerary planning. In order to recognize different traffic signs, we propose that using Zernike moments as the features of traffic signs, and SVM is used to classify the features. Real road scene images are used to evaluate the performance, and five classes of traffic signs are considered. Simulations show that our approach can recognize ground traffic signs effectively and practically.