Abstract: Intelligent transportation system(ITS)  is currently a hot research subject,and the key technology in ITS is traffic signs feature extraction technique.This article,focuse on the characteristics of the traffic signs extraction,proposes use of the scale invariant feature transform algorithm to extract point characteristics of the traffic signs and use of  minimum distance classifier to classify feature vectors.By Matlab,simulation is used to test the results.The results show that the characteristics of the traffic signs can be better detected.

Key words: intelligent transportation system, scale invariant feature transform, traffic sign, feature matching, minimum distance classifier