lane-detection

A video-based guiding system for autonomous car

Running method:

open the main.m and run.

These operations can be chosen according to the specific situation

• set the value of 'pfilename' in main.m to read the specific videoobject
• set the values of variates of 'startno' and 'endno' in main.m to control the first and last frames
• set the related camera parameters and IPM parameters in GetInfo.m

File description:

• main.m: the main function
• GetInfo.m: extract the related parameters of camera and IPM
• dfRoi.m: construct the trapezoid region of interest
• edgedetection.m: edge detection for the IPM grayscale image
• perspectivetrans.m: perspective transform
• TransformImagetoGround.m: transform the video image into the ground coordinate system
• TransformGroundtoImage.m: establish the mapping relationship between pixel coordinate and original image coordinate
• getvanishingpoint.m: get the position of vanishing point to determine the IPM area
• linefit.m: accomplish the lane detection
• turnfit.m: accomplish the lane tracking
• road lane video files: 'challenge.mp4'

Example: