Table of Contents

1. Introduction
2. Approach
3. Dependencies
4. Running the Code
5. Directory Structure

Introduction

This repository contains solution to coding challenge lane detection for self driving cars.

Approach

1. Some basic image processing concepts used here are - Color Selection, RoI Selection, Grayscaling, Gaussian Smoothing.
2. Other concepts used specifically for lane line detection include - Canny Edge Detection, Hough Tranform Line Detection.
3. The notebook uses all of the above techinques to detect lane lines of different types and colors from a video of the road.
4. The pipeline used here - image processing->edge detection and extrapolation->overlay detection on original image.

Dependencies

Python libraries: os, opencv, pandas, numpy, moviepy, imageio, matplotlib

Running the Code

1. Jupyter notebook LaneDetection.ipynb (Python kernel) is the master file.
2. It makes use of:

• solidWhiteCurve.jpg
• challenge.mp4

3. The repository directory structure given below must be maintained for the code to run successfully.

Directory Structure

The directory structure for this repository is as follows:

├── README.md 
├── Scripts
    └── LaneDetection.ipynb