MSER is a subcomponent drawn from the VLFeat open source package (http://VLFeat.org), version 0.9.4.1.
The version of MSER is .5.

The VLFeat.org webpage contains the following text:

http://vlfeat.org/index.html (retrieved September 7, 2009):

<p>The <b>VLFeat</b> <a shape="rect" href="http://www.gnu.org/licenses/old-licenses/gpl-2.0.html">
  open source</a> library implements popular computer vision algorithms  including
  <em>SIFT</em>, <em>MSER</em>, <em>k-means</em>, <em>hierarchical
  k-means</em>, <em>agglomerative information bottleneck, and quick shift</em>.
  It is written in C for efficiency and compatibility, with interfaces in
  MATLAB for ease of use, and detailed documentation throughout. It supports
  <em>Windows</em>, <em>Mac OS X</em>, and <em>Linux</em>. The latest version of
  VLFeat is <code>0.9.4.1</code>.</p>

Thus, MSER is licensed under the GNU General Public License (GPL), version 2.0.

We have included a copy of the GNU GPL 2.0 in this directory, in the file gpl-2.0.html.

We have modified MSER in the following ways:

- we generated a standalone C version from the mser.mex.c source.
- we augmented the code in order to support collection of benchmarking data (matlab and C)
- we create benchmark inputs for the code (matlab and C)
- we modified the code to use our own standardized library for common functions (matlab and C)

Since GPL is viral, our modifications to this benchmark are also open source, under the
GPL 2.0 license.