Skip to content

Latest commit

 

History

History
61 lines (50 loc) · 1.32 KB

File metadata and controls

61 lines (50 loc) · 1.32 KB

Lane detection and turn prediction

Developed a lane detection algorithm using edge detection, Hough transform, homography, and curve-fitting techniques. Achieved 95% accuracy in predicting car turn direction and computing curvature radius. Used semantic segmentation to identify the drivable areas.


Results

  • Lane detection
    • alt text
  • Turn prediction
    • alt text

Dependencies

Plugin
scipy
numpy
cv2
matplotlib
glob

Instructions to run

Problem 1.a) Histogram equaliztion - Full image:

python prob1_all_frames_video.py

or

python3 prob1_all_frames_video.py

Problem 1.b) Adaptive Histogram equaliztion - :

python prob1b_all_frames_video.py

or

python3 prob1b_all_frames_video.py

Problem 2) Lane detection:

python prob2_video.py

or

python3 prob2_video.py

Problem 3) Turn prediction:

python prob3_video.py

or

python3 prob3_video.py

**Find all the outputs after running the code in their respective output folder eg:prob1_output, prob2_output, prob3_output **