hbenameur at yahoo.ca
>,
<Francois.Clement at inria.fr
>,
<Pierre.Weis at inria.fr
>.
|
Ref-image, Image Segmentation by Refinement, is an an image segmentation program using optimal control techniques.
(a) | (b) | (c) |
(d) | (e) | (f) |
Image segmentation using the refinement indicators technique. (a) initial image to be segmented. (b) to (f) first iterations of the algorithm giving segmentations with successively 1, 2, 3, 6 and 10 homogeneous zones. |
The image segmentation problem can be viewed as an optimal control technique: the idea is to consider image segmentation as a parameter estimation problem. The parameter to estimate is the color of the pixels of the image. The adaptive parameterization technique builds iteratively an optimal representation of the parameter into uniform regions that form a partition of the domain, hence corresponding to a segmentation of the image (see Figure above). An error function is minimized during the iterations, and the partition of the image into regions is driven by the gradient of this error. The resulting segmentation algorithm inherits desirable properties from its optimal control origin: soundness, robustness, and flexibility.
The Ref-image user's manual is available on line.
You can also download the manual as DVI, PS, and PDF files.
The manual only exists in the English version, volunteers to translate it are warmly welcome...
You will need OCaml version 4.00 or higher to build Ref-image (this software is tested under Linux and Mac OS X up to OCaml version 4.06).
See the
INSTALL
file provided in the distribution for installation tips.
The latest version (1.2+pl0) of the sources of Ref-image are available via HTTP.
See the
CHANGES
file provided in the distribution.
This program is free software distributed under the 3-clause BSD license.
See the
LICENSE
file provided in the distribution.
If you find a bug in Ref-image, please write to
<refinement-bugs at inria.fr
>.
If you want to get in touch with the implementors of Ref-image
(to contribute some new ideas or lines of code), please write to
<refinement-devel at inria.fr
>.
If you want to get in touch with the users of Ref-image, please write to
<refinement-users at inria.fr
>.
The version 1.2+pl0 of Ref-image was developed from 2009 to 2017 by
hbenameur at yahoo.ca
>,
Francois.Clement at inria.fr
>, andPierre.Weis at inria.fr
>.
The algorithm implemented in Ref-image comes from papers co-written with
Guy Chavent
<Guy.Chavent at inria.fr
>.
Ref-image is entirely written in OCaml.
Page generated by fclement on lamerluche
at 2017-11-23T23:55:32Z.
Latest update of the content of this page:
Thursday, November 23rd, 2017.
Copyright
© 2009-2017
Institut National de Recherche en Informatique et en Automatique (INRIA), all rights reserved.