Research On The Technology Of Traffic Sign Recognition System In Complex Scene

With the expansion of the scale of urban construction,the living standard of the people is gradually improving,drastically speed up the rhythm of people's lives,private cars have become an important means of transport for many people travel.However,as the road motor vehicle continues to increase.At the same time of bring convenience to people,also caused some traffic problem that not allow to ignore-road traffic congestion and safety.And these problems on the one hand and the driver's driving behavior is closely linked,such as drunk driving,road rage,driving experience and non-standard driving behavior;on the other hand related to our country has yet to improve traffic management measures,how to establish a standardized and effective solution has become a common around the world to overcome the difficulties.At present,there are many scholars that: large-scale popular application of intelligent transportation technology,improving the level of road traffic management and the quality of intelligent driving system is an effective way to realize the healthy development of road traffic system.Traffic signs as the basis of the road an important auxiliary facilities,Through the provision of instructions and prohibitions and other driving information for traffic flow guidance,and regulate the driver's traffic behavior.To ensure that the order of traffic orderly manner.Road traffic sign recognition(RTSR)is one of the key technologies of its intelligent driving assistance system,attracting many scholars and relevant scientific research institutions attention and research.For the application of traffic signs in the actual scene identification system,facing the increasingly complex background,and the design of the recognition system to the requirement of real-time and high efficiency is higher,this is undoubtedly a huge challenge.This paper focuses on the key algorithms for the location and identification of road traffic signs:First of all,according to the traffic signs of the shape characteristics,the edge detection method based on wavelet low frequency component is used to extract the edge information of traffic sign image,then,carries on the grayscale morphology and texture noise reduction processing,weakening the noise interference region in the gray edge image,the gray scale image after noise reduction is binarized,later,with the improved Hough transform to segment the target area,and location of traffic sign image.The algorithm of the traffic sign image taken in the natural scene is tested,the experimental results show that the success rate of location is 91.8% and the averagedetection time is 0.46s;Finally,the Hu invariant moments are used to extract the sample and the feature vector of the marker image,and use the support vector machine to classify and identify its meaning,at the same time the minimum distance classification method is compared.