A Lane Departure Warning Algorithm Based On Android Platform Of Research And Implementation

With the rapid development of the automobile industry,convenient transportation also poses a potential safety hazard for people.Traffic accidents occur frequently,mostly due to unsafe driving.In order to reduce traffic accidents and ensure traffic safety,Advanced assisted driving system is one of the current research hotspots.Lane departure warning system is an important part of automotive safety technology,the system will send a warning signal to remind drivers to effectively reduce traffic accidents caused by lane departure when the driver unconsciously deviates from the lane during driving.This article takes the lane departure warning as the research content.,includes parts: image preprocessing,lane line detection,lane line tracking and lane departure warning.The thesis work and research results mainly include the following aspects:1.The algorithm uses the Hough transform to detect the candidate set of lines,and then combines the vanishing point constraint and the RANSAC algorithm to screen the lanes of the candidate to obtain a reasonable lane line,thereby deciding whether the vehicle deviates from the lane.2.Separate left and right lane line parameters are tracked using separate Kalman filters.3.A departure warning algorithm is proposed to estimate the relative lateral offset of the vehicle using the slope of the lane line and its movement,which improves the robustness of the system.4.The lane departure warning algorithm was ported to the Android platform and tested.The test results show that the application of the lane departure warning system implemented on the Android platform has good rate and real-time performance.