21st Signal Processing and Communications Applications Conference (SIU), CYPRUS, 24 - 26 April 2013
Automatic traffic sign detection and recognition is one of the most important components of advanced driver assistance systems. In this paper, a novel method utilizing histograms of oriented gradients based features together with a recently developed and successful circle detection algorithm is proposed for circular traffic sign recognition. In the proposed method, irrelevant backgrounds of traffic signs, whose locations on images are identified within rectangular boundaries, are filtered by detecting their actual circular boundaries. In this way, features representing the traffic signs better can be extracted. The results of the experimental study conducted on a considerably large database demonstrate that the proposed method offers a higher classification performance than the case in which the circle detection is not applied.