The Traffic Sign Detection based on the visual saliency

Abstract

In this paper, we propose a new approach to detect salient traffic signs, which is based on visual saliency and the auto segmentation method based on region merging. The proposed algorithm introduces the visual saliency model, and puts forward the top-down and bottom-up two-way fusion mechanism, tried to consciously take the initiative to find and simulate human visual unconscious passive visual process by attracting interact. Then the segmentation method based on region merging is used to extract the traffic sign. Experiments show that the two-way fusion mechanism in target detection has fast processing speed and high accuracy. Extensive experiments on public datasets show that our approach outperforms state-of-the-art methods remarkably in salient traffic sign detection. Moreover, the proposed detection method has higher accurate rate and robustness to different natural scenes.