Skip to content

Latest commit

 

History

History
41 lines (31 loc) · 1.65 KB

README.md

File metadata and controls

41 lines (31 loc) · 1.65 KB
Raw

Use OpenCV to create crosshatch-style drawings and videos.

Try out a canvas-based implementation at https://www.sketchify.me

To use:

$ make
$ ./crosshatched sources/goldengate.jpg
$ ./crosshatched --laplacian sources/thailand.jpg
$ ./crosshatched path/to/a/video.mp4

This script is optimized for images and videos >= 1080 pixels wide. You can also get decent results out of smaller images by tweaking the constants although bigger seems to be better in most cases.

There are a bunch more flags too worth playing around with

The algorithm:

  • Calculate the gradient of the image
  • Draw short, connected bezier lines parallel and perpendicular to that gradient
  • Generate the "edge gradient" from either the Laplacian or Canny (default) algorithms
  • Using slightly tweaked rules, draw more bezier lines parallel to the edge gradient

Dependencies:

  • g++ (Part of the GNU Compiler Collection)
  • OpenCV
  • Tons on other stuff (install everything one error at a time)

TODO:

I'd love to pipe this through a neural-network style-transfer first to get something really artistic

A note:

I'm generally a Python programmer so please forgive my shitty C++ code. I'm also not at work so I'm not keeping my code clean