Advanced Lane Finding for Self-Driving Cars

Lane detection for self-driving cars

To perform lane finding for self-driving cars, we can outline some goals as the following:

  • Compute the camera calibration matrix and distortion coefficients given a set of chessboard images.
  • Apply the distortion correction to the raw image.
  • Use color transforms, gradients, etc., to create a thresholded binary image.
  • Apply a perspective transform to rectify binary image (“birds-eye view”).
  • Detect lane pixels and fit to find lane boundary.
  • Determine curvature of the lane and vehicle position with respect to center.
  • Warp the detected lane boundaries back onto the original image.
  • Output visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.

First, we compute the camera calibration matrix and distortion coefficients given a set of chessboard images.

Second, we will run through our image pipeline, starting with our original image.

We need to apply a distortion correction to the raw image. Notice how the street sign is straight now.

Then we use a color transforms, gradients, etc., to create a thresholded binary image.

We obtain the lanes from the previous image and then apply a perspective transform to rectify binary image and create a “birds-eye” view.

Now we can detect the lane pixels and fit to find the lane boundary.

With the lane boundary detected, we can determine curvature of the lane and vehicle position with respect to center and then warp the detected lane boundaries back onto the original image.

On the image, we can output the visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.

Avatar
Eric M. Fischer
MS Artificial Intelligence

My research interests are in generative modeling for computer vision and natural language processing. Specifically, I am interested in applying generative learning techniques to language problems.

Related