Lane Detection for Autonomous Vehciles

This is the python implementation of lane detection used in autonomous vehicles. The techniques used in this work are geometric vision based and are not robust for all scenarios.

(https://www.youtube.com/watch?v=E52yUwqryhY&ab_channel=SandeepKota)

(https://www.youtube.com/watch?v=CTZXtOfOYFI&ab_channel=SandeepKota)

Dependencies

This code was tested on a system with following packages:

• Python - 3.6
• OpenCV - 3.4.5
• Numpy
• Matplotlib

Directory Tree

The directory list for the dataset is given below

.
├── code2.py
├── code.py
├── Data1.mp4
├── Data2.mp4
├── Dataset
    ├── data_1
    │   ├── camera_params.yaml
    │   └── data
    │       ├── 0000000000.png
    │       ├── 0000000001.png
    └── data_2
        ├── cam_params.yaml
        └── challenge_video.mp4

Working

Instructions to run the code for Improving Video Quality

• The python script improve_video_quality_ques1.py runs on every individual frame of the video.

python3 improve_video_quality_ques1.py ../Data/Night\ Drive\ -\ 2689_Problem1.mp4

Instructions to run the code for Lane Detection

• The python script code.py is the lane detection script for the dataset 1. To run the code, open a new terminal in the directory and type

python3 code.py

• The python script code1.py is the lane detection script for the dataset 2. To run the code, open a new terminal in the directory and type

python3 code2.py