Research On The Recognition Technology Of Traffic Signals Based On Machine Vision

With the rapid economic development,people's living standards gradually increased,cars are becoming more and more popular.However,the car brings convenience to people.But at the same time,it also makes the road traffic carrying pressure growing.In order to alleviate the traffic pressure and reduce the huge losses caused by traffic accidents,the development of intelligent driving assistance system which can extract the key traffic information in real time and make timely warning becomes the research focus.Automatic detection based on red light signal as the research target,based on machine vision theory and image processing technology to drive in a real scenario for image processing.The main work and outstanding contributions are as follows:(1)The selection method of red light is proposed based on RGB and HSV.Obtaining the threshold of two color spaces through the experimental analysis.If the color stays within the threshold range,then the image segmentation is then used to extract the suspected target area.(2)According to the geometry,position,and relation of the traffic light,a kind of screening method for the various kinds of geometry,color and position characteristics of red light is proposed.In suspected target,the use of the geometry of a red light,color,location features we don't need to delete the target area of impurities,retain the red light area,complete intelligent extraction of red light information.(3)The overall implementation and technical route of the automatic detection of red light signals will be designed and integrated into a smart mobile terminal based on the android platform.In 15000 images under different natural environment and conditions of the experiment test,the average speed of 60 ms,fully meet the real-time requirements,to each kind of environment comprehensive test success rate reached 90%.In this paper,for intelligent mobile terminal in intelligent auxiliary driving system provides a feasible way,and for the extraction of the traffic information intelligent further research provides a theoretical basis and experimental data.