Lane Detection in RPi for a RC rover or Self Driving Vehicle using Pi Camera. This project uses a CNN model. This is an extension of the MLND Captsone Project that uses a similar approach. This project is a part of the autonomous triwheeler college project.
Please let me know any implementable solutions of interest for the triwheeler if you have any.
The dataset includes 1280px X 720px video from smartphone which is scaled down to 360 px x 120 px. And labels were assigned containing two lines. Having 3 parameters for each line. 3 times 2, making it a 6 parameterized labels.
pip install keras
Download openCV from https://packages.debian.org/buster/armhf/python3-opencv/download Navigate to the directory and type the following on the terminal.
dpkg -i python3-opencv_3.2.0+dfsg-4+b4_armhf.deb
Attach the Pi Camera to th Rpi3 board and run the followingg command.
python3 draw_detected_lanes.py
- Yash Gupta