Skip to content

Latest commit

 

History

History
104 lines (63 loc) · 3.45 KB

writeup_template.md

File metadata and controls

104 lines (63 loc) · 3.45 KB
Raw

Finding Lane Lines on the Road

Writeup Template

You can use this file as a template for your writeup if you want to submit it as a markdown file. But feel free to use some other method and submit a pdf if you prefer.

Finding Lane Lines on the Road

The goals / steps of this project are the following:

  • Make a pipeline that finds lane lines on the road
  • Reflect on your work in a written report

Reflection

1. Describe your pipeline. As part of the description, explain how you modified the draw_lines() function.

My pipeline consisted of 5 steps.

  1. I converted the images to grayscale.

  2. I applied Gaussian smoothing. The kernel size is 5.

  3. I applied Canny edge detector with the following parameters.

    • low threshold = 50
    • high threshold = 150

  4. I masked the images.

    • vertices = (bottom left, (440, 320), (520, 320), bottom right)

  5. I applied the Hough transformation with the following parameters.

    • rho = 1
    • theta = np.pi/180
    • threshold = 10
    • min line length = 40
    • max line gap = 20

In order to draw a single line on the left and right lanes, I modified the draw_lines() function by the following procedures.

  1. I separated line segments to left or right lane by their slope.

    • left lane: tan(-pi/3) < slope < tan(-pi/9)
    • right lane: tan(pi/9) < slope < tan(pi/3)
    • dump: else

  2. I calculated the slope of the left and right lanes by averaging the segments slopes categorized into them.

  3. I calculated the intersection point of the each lanes and y = 320 by averaging the intersection points of the each line segments and y = 320.

  4. I calculated the intersection point of the each lanes and the bottom line.

  5. I drew a single line on the left and right lanes.

If you'd like to include images to show how the pipeline works, here is how to include an image: Put your images into the test_images folder.

2. Identify potential shortcomings with your current pipeline

  1. When resolution of images are changed, my pipeline would get wrong line lanes because mask and finding lane process is rely on horizon line defined y = 320.

  2. When road surface is wetty and reflects lights of signals and street lights, my pipeline would get wrong line because it can't distinguish between edges from lanes and lights.

  3. When the car turn, my pipeline would get wrong line lanes because my pipeline distinguish the left lane and the right lane by slopes of line segments.

3. Suggest possible improvements to your pipeline

A possible improvement for shortcoming no.1 would be to decide horizon line by the proportion of the height of image. (I assume that the angle of the camera is constant)

A possible improvement for shortcoming no.2 and no.3 would be to identify the left and right lanes by not only the slopes of line segments but the positional relationship with the car.