--------------------------------------------------------------------------------
Description
--------------------------------------------------------------------------------
This  directory  contains the implementation of the salient region detector des-
cribed in the paper:

    Joint upsampling of random color distance maps for fast salient region
    detection
    Maiko M. I. Lie, Gustavo B. Borba, Hugo Vieira Neto, Humberto R. Gamba
    Pattern Recognition Letters, 2017
    http://dx.doi.org/10.1016/j.patrec.2017.09.010

The source code is written in  MATLAB,  and makes  use of third-party  C++  code 
using a MEX interface. If you use this source code,  please cite the work above.
Access the DOI link above for up-to-date reference information.

A previous version of this method was published in a conference paper:

   Fast Saliency Detection Using Random Color Samples and Joint Upsampling
   Maiko M. I. Lie, Gustavo B. Borba, Hugo Vieira Neto, Humberto R. Gamba
   Proceedings of the 29th Conference on Graphics Patterns and Images, 2016 
   http://dx.doi.org/10.1109/SIBGRAPI.2016.038

--------------------------------------------------------------------------------
Files
--------------------------------------------------------------------------------
The following files should be in this directory:

    JUSAL.m     Salient region detector function.
    mexFGS.cpp  Fast Global Smoother (FGS) MEX file, from Min et al. (2014).
    compile.m   Compilation script for the MEX file.
    README      This file.

--------------------------------------------------------------------------------
Usage
--------------------------------------------------------------------------------
First,  you will need to compile the MEX file. Within  this directory,  or after 
adding it with addpath(), run:

    compile

(your MATLAB  environment must have a C++  compiler properly configured. If this
is not the case, refer to the MathWorks documentation for how to do this)

For a demo, run:

    I = imread('peppers.png');
    S = JUSAL(I, 0.25, 0.2, 3);
    figure; imshow(I); figure; imshow(S)

For more information, including a description of the arguments, run:

    help JUSAL

--------------------------------------------------------------------------------
Third party source code
--------------------------------------------------------------------------------
This  work uses the  Fast Global Smoother (FGS) by Min et al. (2014),  including
their source code, which can be accessed in the following link:

    https://sites.google.com/site/globalsmoothing/

The reference for their method is:

    Fast Global Image Smoothing based on Weighted Least Squares
    D. Min, S. Choi, J. Lu, B. Ham, K. Sohn, and M. N. Do,
    IEEE Trans. Image Processing, vol. no. pp., 2014.

If you use the parts of our work that include this source code, please also cre-
dit their work. In this case, we also recommend that you read their own README
and license information.

                                                         Maiko Lie
                                                         Last updated: Oct, 2017