Details of project : [ Ссылка ]

Advanced Lane detection project in Udacity course:
The project involves following steps :
1) Converting the distorted image to undistorted image(Distortion could be due to camera position or type of lens)
2) Converting the image into binary image using various transformation(HLS or HSV)
3) Applying the perspective transformation (often called Bird's eye view which is looking from top of car). This is the interesting topic.
4)Then detect the lane boundary by applying the histogram and giving a threshold for left and right lanes.
5) Determining the lane curvature (very interesting topic) with respect to center. In video we see it as 'R' and 'pos'
6)Once we do this in the transformed image, we go back by reverse transformation.
7) Coloring the identified boundary by some maths and programming! ;)


This project is better than canny edge detection for lane detection because of the fact that, lightness variations doesnt affect in lane detection here ;)

свернуть