Vision-Based Stop Sign Detection and Recognition System for Intelligent Vehicles

Abstract

The traffic sign detection and recognition system is an essential module of the driver warning and assistance system. A vision-based stop sign detection and recognition system is presented here. This system has two main modules: detection and recognition. In the detection module, the color thresholding in hue, saturation, and value color space is used to segment the image. The features of the traffic sign are investigated and used to detect potential objects. For the recognition module, one neural network is trained to perform the classification and another one is trained to perform the validation. Joint use of classification and validation networks can reduce the rate of false positives. The reliability demonstrated by the proposed algorithm suggests that this system could be a part of an integrated driver warning and assistance system based on computer vision technologies.