This package provides standard functions to create and manipulate edge-weighted graphs of images: create a graph, segment it, compute its watershed, or its connected components...
1/ Torch7 is required:
Dependencies, on Linux (Ubuntu > 9.04):
$ apt-get install gcc g++ git libreadline5-dev cmake wget libqt4-core libqt4-gui libqt4-dev libboost-all-dev
Dependencies, on Mac OS (Leopard, or more), using Homebrew:
$ brew install git readline cmake wget qt
Then on both platforms:
$ git clone https://github.com/andresy/torch
$ cd torch
$ mkdir build; cd build
$ cmake ..
$ make
$ [sudo] make install
2/ Once Torch7 is available, install this package:
$ [sudo] torch-rocks install imgraph
First run torch, and load imgraph:
$ torch
> require 'imgraph'
Once loaded, tab-completion will help you navigate through the library:
> imgraph. + TAB
imgraph.colorize( imgraph.connectcomponents(
imgraph.graph( imgraph.histpooling(
imgraph.segmentmst( imgraph.testme(
imgraph.watershed( imgraph.gradient(
To get quickly started, run the testme() function:
> imgraph.testme()
which computes a few things on the famous image of Lena: