Based on Image Processing using Deep Learning + OpenCV
- A project based on Image Processing using Convolutional Neural Networks + OpenCV.
- Pi CAM was used for all the real time image input.
- Images were simplified using OpenCV i.e hough transform, region masking etc to make it easier for our neural network to identify features.
- Raspberry Pi 3 was used for nominal real time image processing using open CV.
- The model was trained using KERAS library with 2000 images each for "left, right, straight, sharp left, sharp right" to follow a track.
- The motors were controlled using Arduio UNO with serial communication.
- This robot could also be used as a remote control car to race around the track. ;)