Skip to content

Efficient graph-based image segmentation procedure with OpenCV+igraph

Notifications You must be signed in to change notification settings

dnery/DIP-project

Repository files navigation

You'll need to have installed on your system:

  • OpenCV (C++ headers & python3 package)
  • igraph (C package)
  • CMake >= 2.6

Compile:

$ mkdir build
$ cd build
$ cmake ..
$ make

Usage:

$ ./image_to_graph <image name> <pixel neighbourhood radius> <pixel similarity constant> [reduce: 1..3 (default: 1)]

The docs folder contains the project documentation, including the presentation, experiment analysis and final considerations. The run_tests.sh script will run the program on some of the input images in the assets folder and output several segmentation results in the results folder. The experiment result filenames are given as such:

<input image name>.<pixel neighbourhodd radius>_<pixel similarity constant>_<OpenCV downscaling factor enum>.<input image format>

About

Efficient graph-based image segmentation procedure with OpenCV+igraph

Resources

Stars

Watchers

Forks

Packages

No packages published