Traffic signs recognition based on visual attention mechanism

Abstract In order to improve the speed and accuracy of traffic signs recognition (TSR), a novel method based on visual attention mechanism is present in this paper. According to this method, the potential regions of traffic signs could be detected by extracting color contrast features in the CIELab color space based on visual attention model, meanwhile dynamic thresholds are used to adjust the timing of the inhibition of return. Then scale invariant feature transform (SIFT) feature is extracted for the candidate region, the similarity calculation is done between the candidate region and the standard traffic signs, thus, the traffic sign would be recognized. Compared with traditional methods, the proposed method can achieve rapid and accurate detection and recognition and has the characteristics of no need for segmentation, less calculation, and bionics. The experiments results on two traffic signs databases collected from domestic and Netherlands show the method's excellent effect on TSR.