Road Traffic Sign Identification in Weak Illumination for Intelligent Vehicle Based on Machine Vision

摘要

Background: For life, cars have brought convenience to people, besides causingenvironmental pollution, traffic congestion, and road safety. One of the main reasons for traffic accidentsis that drivers cannot respond to road state and information timely. Intelligent vehicle and safetydriving assistance technology are one of the ways to reduce traffic accidents. Many papers and patentsare studying how to reduce the rate of traffic accidents.Objective: Traffic sign recognition is an important part of the intelligent vehicle system and advanceddriver assistance system. To reduce the rate of traffic accidents in weak illumination, this paperproposes a detection and recognition method for intelligent vehicle traffic sign based on machine vision.Methods: Firstly, the traffic signs are reclassified from the color and shape characteristics of trafficsigns. Then, the concept of color shape pairs is put forward, and the color-geometric model is established.To achieve the enhancement effect, the image is preprocessed by the uniformity of the histogram.We mainly use the fast algorithm of invariant moments and Zernike moments to extract the signfeatures. Finally, template matching and support vector machine are used to traffic sign image recognition.Results: The experimental results show that the technology has good robustness in complex environmentssuch as weak light, occlusion and shadow and it can improve the image recognition rate effectively.Conclusion: Compared with the traditional method, it provides a method for intelligent vehicle trafficsign identification under weak illumination conditions.