Detection and Extraction of Road Traffic Signs

Abstract

The real-time road traffic sign recognition focused on the application of computer vision techniques in automatic detection and classification of video images of traffic signs placed on roadways acquired from moving vehicles. In this work, we present results of an attempt to locating traffic signs placed on local roadways based on image processing techniques. The developed algorithm is based on two basic processing stages; localization and extraction of the traffic signs. The algorithm takes a raster image of a background scene with a traffic sign as the input and extracts the sign as the output. The preliminary work shows high performance in the extraction of traffic signs with warning messages on local roads. The algorithm was tested for a variety of images and showed 98% accuracy in locating and extracting traffic signs.