Skip to content

In this project, I developed a pipeline on a series of individual images, and later applied the result to a set of video streams.

License

Notifications You must be signed in to change notification settings

Arina-W/Finding-Lane-Lines

Repository files navigation

Self-Driving Car Nanodegree : Finding Lane Lines on the Road

Writeup Report : Project 1/9

In this project, I developed a pipeline on a series of individual images, and later applied the result to a set of video streams.


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 the process in a written report (This report)

Reflection

1. Pipeline descriptions and the draw_lines() function.

My pipeline consists of 5 steps I stated down below.

  1. Color Detection : Images are converted to grayscale using openCV library.
  2. Blurring : Gray images are blurred using the Gaussian Blur function.
  3. Finding edges : Blurred images are then used to detect edges using Canny function, also openCV.
  4. Region Masking : A mask is created to define ROI of image using numpy libary.
  5. Using Hough Transform function, Canny images is taken in and return an image with hough lines drawn on it.

In step 5), in order to draw a single line on the left and right lanes, draw_lines() function is modified. Modification consists of separating line segments by their slope to decide which segments are part of the left line vs the right line. This function will draw lines with parametes for color, and thickness. Lines are drawn on the image.

To better visualize the result, here is an example of an image output:

output

2. Potential shortcomings with the current pipeline

Current pipeline might have trouble to perceive the lane lines if the car needs to take a sharp turn or even a simple curve maneuver.

Another concern would be the effect of nightlight from night lamps and reflections of headlamps etc during nighttime.

3. Possible suggestions for improvements to the pipeline

A possible improvement would be to also apply small segmentations of the image to acknowledge even the sharpest turn or maneuver for the car.

To cater problems that might occur at dark, I suggest that we use value scale in the HSV coordination and apply a mask of it in the pipeline.

About

In this project, I developed a pipeline on a series of individual images, and later applied the result to a set of video streams.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published