Considered as the starting point for a Self-Driving Engineer due to its low complexity, this project will show an approach to find the lane boundaries on a video taken from the roof of a vehicle in the highway.
This repository explains the approach implemented for the project number 1 of the Udacity Self-Driving Car Engineer Nanodegree.
In this repository an algorithm based on pure computer vision techniques is shown and explained in order to provide a great guide for the reader.
The tools used to find the lane lines in the images and the videos were:
- Python
- OpenCV
Access the Jupyter Notebook of the project. In it, the approach is explained in depth and the result for each step can be seen.
The approach used in this project consists in seven steps:
- Convert the RGB input image into a Grayscale image. See cell
4 - Smooth the Grayscale image with a Gaussian Blur filter to the Grayscale image. See cell
5 - Apply the Canny edge detector to the smoothed image. See cell
6 - Select and extract a region of interest (ROI) from the image. See cell
7 - Apply the Hough transform to find the lines and its points within the ROI. See cell
8 - Group the lines based on their slope to find one that represents all of them. See cell
9 - Combine the input RGB image and the ROI image which contains the lines to display the detections. See cell
10
Finally, in order to process each frame of the video and obtain the final output, a function that blends all of steps explained above is created. See cell 11
Find a video with the final results of this project in this link.