Project: Finding Lane Lines on the Road

Overview

The project contains two methods for finding Road lines.

The Basic pipeline detects the road line segments, then extrapolate them and draw them onto the image for display. Within the pipeline some of the tools used are: color selection, region of interest selection, grayscaling, Gaussian smoothing, Canny Edge Detection and Hough Transform line detection.

The advance pipeline contains a series of functions using polynomial fits, and splines that can help identify, draw and Measure Curvature of the road lane lines from images and video.

For more information about this project visit the Wiki page

Quickstart

Download repository To run the program on the existing images under "test_images". Open command line under the project directory and type the following.

python main.py

Note: You can replace the images found in "test_images" with your own road images

By default the the pipeline is advance, you can specify the pipeline you would like to use.The advance pipeline can be called by setting argument one 2 in the command line:

python main.py 2

The basic pipeline can be called by setting argument one 1 in the command line:

python main.py 1

(Optional) If you would like to process videos, all you have to do is add to the command line an argument with the directory where the videos are located.Example:

python main.py 2 ../test_video/

The output video or images will be stored under test_results directory

Dependencies

Python 3.5, openCV 3.2

References

This project is part of the CarND program. Tools learned in class were used to identify lane lines on the road.