Lane lines are a constant reference for where to steer the vehicle. One of the first things we would like to do while developing a self-driving car is to automatically detect lane lines. The goal of this project is build a lane detector to detect lane lines on the road.
- Convert image to grayscale
- Blur the image using Gaussian blur
- Use Canny edge detector to get edges of the image
- Blackout everything outside the region of interest
- Use Hough transform to detect lines in the edge-world image
- Draw the detected lines in the original image
- I have not supported different lighting conditions. Any dark images would probably screw up the lane predictions.
- The Hough transform is currently only looking for straight lines. Unless you live on a chess board, this is not good enough.
- Some roads may not have lane lines at regualr intervals. I need to sort this out.