Real-Time prototype of driver assistance system for Indian road signs

Abstract

The paper presents a system for the detection and recognition of road signs in real time (tested for Indian road signs). The detection and recognition algorithm used is invariant to scale, angle, blur extent, and variation in lighting condition. Shape classification of road signs using Hu moments is done in order to categorize signs as either warning, mandatory, prohibitory or informational. Classified road signs are then matched to ideal road signs using feature extraction, and the matching is done with the help of Oriented FAST and Rotated BRIEF (ORB) descriptors. After recognition, the driver is given a feedback.