Advanced Lane Finding (ALF)

This is a Udacity Self-Driving Car NanoDegree project submission that uses the following to detect road lanes:

• Lens calibration
• Distortion correction
• Edge detection
• Colorspace Transformation
• Perspective transformation
• Pixel dection and Line Fitting

Installation

Clone or fork this repository.

Usage

Intended user is the Udacity evaluator for this project. It is intended to be used in a Jupyter Notebook. Open alf.ipynb for examples of usage.

Files

Project Files

• writeup.md: writeup of project for Udacity evaluator; includes images of pipeline stages
• output_images/project_video.mp4: project video for submission
• alf_*.py files: python code for project
  • alf_con.py: main controller for sequencing pipeline stages
  • alf_cam.py: camera for calibration and distortion correction
  • alf_enh.py: enhancer for edge detection and HSV color transformation
  • alf_war.py: warper for perpective transformation
  • alf_llg.py: lane finders, sliding and linear window search areas, and line to find and annotate lanes
  • alf_hud.py: heads up display (simulated) for composing final image with lane area
  • alf_utils.py: logging and demonstration functions

Other files

• alf.ipynb: jupyter notebook on project usage and demonstrations of important functions
• output_images folder:
  • wup_*.png: images used in writeup
  • project_video.mp4: project video for submission
  • challenge_video.pm4: pipeline worked on challenge video too
  • harder_challenge_video.mp4: pipeline failed on this harder challenge video
  • out_*_video_.mp4: output of videos processed in alf.ipynb examples; prevents overwriting project videos for submission
  • straight_lines1-undist.jpg: image used to find and plot srcpoints for perspective transformation
• sandbox.ipynb: jupyter notebook for testing code
• sketch.drawio: UMLish sketch of component collaborations; requires diagrams.net to view; does not reflect current state of components
• adv_lane_fine.log: debug log of project