Skip to content

Estimate and visualizes an ego-vehicle’s trajectory in the ground frame, using a traffic light as a world reference point. Demo code for Wisconsin Autonomous.

Notifications You must be signed in to change notification settings

li21rich/Computer-Vision-Ego-Vehicle-Trajectory

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Perception Code

Methodology - trajector.py

  • Get the center pixel values at every frame of the bounding box for the traffic light
  • I initialized frame 1 first for clarity as a starting point, which allows me to use the offset values for ground frame view.
  • I then iterated through every frame while ignoring noise/NAN/0,0,0 items at O(n) with frames in reverse order
  • Used matplotlib to graph and animate

Assumptions

  • World/ground frame (BEV) sets traffic light at origin
  • Camera coord system (X forward, Y right axis, Z upward)
  • Right-handed coord system for ground frame (X forward, Y left, Z upward)

Results - trajectory.png, trajectory.mp4

  • trajectory.png with a curve that begins somewhere around -27,16 and ends at 0,0 (all meters) computed at O(n)
  • trajectory.mp4 to animate the pathway taken by the vehicle computed at O(n^2)
  • Mapped out traffic light and ego only

About

Estimate and visualizes an ego-vehicle’s trajectory in the ground frame, using a traffic light as a world reference point. Demo code for Wisconsin Autonomous.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages