Simulation And Implementation Of Lane Departure Warning Method Based On Image Processing

With the rapid development of the transportation industry in China,the total number of vehicles in China is growing rapidly,and the problem of traffic safety is becoming more and more serious.The problem of vehicle driving safety is also getting more and more attention from the whole society.In recent years,the traffic accidents in China have occurred frequently,so the research and development of a product which can cooperate with the driver to ensure the safety of the road traffic is getting more and more people's expectation.In order to improve the driver's safety environment in the course of driving,a large number of experts and scholars began to develop an auxiliary driving system to improve the driving environment,and the concept of lane departure early warning system has been proposed.The design aim of the system is that when the driver has been driving for a long time,when the body is overtired and the attention is not focused,the vehicle causes the vehicle to deviate from the front Lane unconsciously,thus giving warning information,prompting the driver to correct the vehicle track in time,preventing the major traffic accidents and ensuring the safety of the personal property.Based on the image processing technology,this paper carries out theoretical research and experimental simulation on the lane detection algorithm,tracking algorithm and deviation early warning decision algorithm used in the system.This paper first expounds the background and significance of the research on lane departure warning system and summarizes the research work of experts and scholars at home and abroad in the lane departure warning system.In order to extract the feature of lane line more clearly,different preprocessing algorithms are used to compare and analyze the image.Finally,the weighted average method is used to carry out grayscale processing.The median filter is used to remove the image noise,and the improved Sobel operator is used to enhance the edge filtering.Then,a lane line model is established based on prior knowledge of the road,and an improved Hough transform algorithm is used to extract the lane line.Kalman filtering algorithm is used to estimate the lane line parameters,and the lane line dynamic ROI area is established.The ROI is applied by least squares method.The edge feature points in the area are subjected to a fitting process to extract the relevant parameters of the lane line.Finally,the lane line information is compared with the vehicle position information to establish a lane departure warning model.The model determines whether the driving vehicle has a lane departure by calculating the distance between the vehicle center point position and the lane boundary line.The algorithm has a small calculation amount.Alarm accuracy is high.This article uses the Matlab2016 and Open CV in Window7 environment to design a lane departure warning system.The effectiveness and accuracy of the system have been verified through multiple detections of lane lines and departure warnings.