Research On Speed Limit Traffic Sign Recognition Based On Machine Vision Technology

With the development of social economy and the improvement of people’s livingstandards, the number of cars is constantly increasing, the car has broughtconvenience to people’s life, but at the same time, a lot of traffic problems also becreated, such as traffic jam, road safety, etc. In order to solve these problems, theresearch field of Intelligent Transportation System came into being. IntelligentTransportation System combines image processing, digital signal processing, artificialintelligence, pattern recognition and communication technology, etc, can effectivelycoordinate the relations among pedestrians, vehicles and roads, ITS is significant toensure the safe operation of transportation system.Traffic sign recognition is an important part of ITS, it can identify the trafficsigns in the driving environment in real-time and accurately in the process of themotion of the car, and the traffic sign information is passed to the driver in time, tostandardize the driver’s behavior, help them to drive safely, to reduce traffic accidentsas far as possible.This paper chooses the speed limit signs as the research object, the identificationtechnology has carried on the thorough research, the main research contents include:1. The detection of speed limit signs: According to the color characteristics ofthe speed limit signs, the paper respectively from the RGB color space and HSV colorspace for color segmentation, the target area separated from the background, thencombined with the round feature of speed limit signs, further confirm the position ofthe speed limit signs in the picture, so as to realize the detection of speed limit signs.2. The recognition of speed limit signs: Recognition technology is the mostimportant work in this paper. By analyzing and comparing the template matchingmethod with the recognition algorithm based on structure characteristics for the speedlimit signs recognition accuracy through experiments, finally decided to adopt themethod based on structural characteristics to recognize the speed limit signs.Finally, this paper evaluates the speed limit signs recognition system, the resultshows that the algorithm has good real-time performance, accuracy and robustness.