Cycle lane segmentation under different light conditions for autonomous robot navigation .

The main task is to try to write a script to segment the cycling lane under different lighting conditions (including shadows) to use the segmented path to control the robot motor. The segmentation should be done by canny edge detection segmentation or colour segmentation. Another option would be using light CNN that can be implemented in the raspberry pi 4 to segment the lane. The script should be in python language. An example of the expected results is given below. The script should work on live streams and pre-recorded videos. A video of the cycling lane will be sent later due to its large size.