Skip to content

Latest commit

 

History

History
72 lines (61 loc) · 2.35 KB

README.md

File metadata and controls

72 lines (61 loc) · 2.35 KB

Video Enhancing, Lane Detection and Turn Prediction

Authors

Prasheel Renkuntla Shubham Sonawane Raj Prakash Shinde

Description

The project demonstrates Video enhancing and Lane detection for self driving cars on the KITTI and Udacity dataset.

Dependencies

  • Ubuntu 16
  • Python 3.7
  • OpenCV 4.2
  • Numpy
  • copy
  • sys
  • argparse

Run

To run the Video enhancement on Night ride video

python3.7 P1-video_correction.py

To run the lane detection on images, 2(a) Udacity Dataset -

python3.7 P2-1-Homography.py

To run the lane detection on challenge video, 2(b) KITTI Dataset -

python3.7 P2-2-challenge_accepted.py

Demo

The original night ride frame is shown on the left and the CLAHE processed frame is on the right for the image given below-

Comparison of original and enhanced video output

For the Udacity Dataset, we have determined the homography points from the image as shown below-

Homography points on the image

The final output for turn prediction and lane detection is given below-

Udacity Dataset

The output from the challenge video is shown here-

KITTI Dataset

The Output folder consists of video results from the KITTI and Udacity datasets only.

Reference