Research On Traffic Sign Detection And Recognition Technology

With the increase in the number of cars,the city and high-speed road conditions are becoming more and more complex,in this context ITS(Intelligent Transportation System)was born.Intelligent transportation system is a system which can effectively fuse various kinds of advanced information technology.In recent years,intelligent transportation system has utilized some of the more advanced image processing technology and pattern recognition technology,so the intelligent transportation system is becoming more and more intelligent.One of the main research areas is the detection and identification of road traffic signs.The video capture device installed in a moving car and take some traffic sign detection and recognition algorithm,can effectively guide the driver specification driving,also can timely pavement feedback to the driver,which can reduce the occurrence of traffic accidents due to driver distraction or distraction.Therefore,the detection and recognition of traffic signs have a very important and broad theoretical research value and practical value,and now for the real environment of traffic signs image detection and recognition is becoming more and more important.However,because of the traffic signs detection and recognition is for real-time applications,and the current research algorithms are not very good to meet the conditions,so there are still many problems in the practical application,the actual effect is not good.In this paper,the original image is processed by image enhancement and image restoration technology,and the quality of the original image is improved.Because of the natural environment of the road traffic signs image is likely to have a light,contrast,image details,such as the problem,so the use of histogram equalization based on traffic signs image enhancement processing and image denoising.Secondly,according to the characteristics of the external environment and the characteristics of different color space,this paper chooses the HSV color space,because the HSV color space is not very sensitive to illumination changes,so the color segmentation of the image is carried out in this space.Once again,for a picture of the possible existence of color and traffic signs similar to the region of this situation,the use of morphological and other operations on the suspected area of the pre treatment,while the use of shapecharacteristics of traffic signs in the area of accurate positioning and cutting down.Finally,this paper uses the template matching based traffic signs recognition algorithm,in order to ensure the real-time performance of the system and greatly improve the accuracy of the identification system.