Speed limit sign board detection and extraction of digits in different weather conditions

Abstract

The goal of this paper is to explore how to find the prohibitory speed detection board automatically by doing image processing. After detection, it detects the numerical portion from the board by OCR operation and informs the vehicles about the maximum speed allowed for that particular road. In this paper, we present a new modular traffic signs recognition system, successfully applied to all part of the world including country in Asia as well as Europe. Our sign detection step is based only on shape-detection (circle). We try to extract the red portion from the image then using Hough transformation (HT) we detect the circle from the image avoiding all other unnecessary information from the image. This system able to detect board in all situation like in bad lighting, noisy images, blurred images, dawn and dusk and also in different environment like foggy condition, snow fall area, sunny bright images with a very high detection rate. Speed sign candidates are classified by segmenting potential digits and then applying neural digit recognition.