Distortion invariant road sign detection

This paper proposes an universal method to detect and recognize road signs of any countries with any color or any of the existing shapes (e.g. circular, rectangular, triangular, pentagonal and octagonal). The presented system is invariant to transformation (e.g. translation, rotation, scale and occlusion). There are three main stages in the proposed algorithm: 1) segmentation based on the color features to find the the region of interests (ROIs), 2) traffic sign detection by using two novel shape classification criteria, and 3) recognition of the road sign using distortion invariant fringe-adjusted joint transform correlation (FJTC) for matching the unknown signs with the known reference road signs stored in the database. Experimental results on real life images demonstrate that the proposed framework is invariant to translation, rotation, scale and partial occlusions.