Skip to content

eshibusawa/Stixel-World-Python

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Stixel World (Stixel-World-Python)

Stixel-World-Python is a python implementation for GPU-accelerated real-time stixel computation [1, 2].

Stixel computation result

left image, stixel image

Differences from official implementation

For convenience of parameter tuning, trial and error, the author implemented the method using CUDA and CuPy. Using of NVRTC runtime compiler has some advantages over AOT compilation, e.g., CMake is not required, easy debugging, visualization and so forth. The implementation has the following differences from official implementation [2].

  • prefix-sum and histogram operation are replaced an implementation using CUB

Requirements

VFS-Python requires the following libraries:

  • cupy
  • opencv

Usage

python main_stixel_world.py

References

  1. Hernandez-Juarez, D., Espinosa, A., Moure, J. C., Vázquez, D., & López, A. M. (2017, March). GPU-accelerated real-time stixel computation. In 2017 IEEE Winter Conference on Applications of Computer Vision (WACV) (pp. 1054-1062). IEEE.
  2. https://github.com/dhernandez0/stixels

Releases

No releases published

Packages

No packages published