Traffic Sign Detection Based on Shape Model

摘要

As an important part of the Intelligent Traffic System (ITS), traffic sign detection is an important pre-procession for ITS. This paper developed a novel sign detection algorithm based on the Shape-band template matching which models an object within a bandwidth of its sketch/contour. Moreover, shape-band is also used as a feature selection method in this paper. Experimental results show the traffic sign detection is robust towards deformation of signs, illumination change and partial occlusion.