View source for Iris Segmentation Code Based on the GST


{{Infobox
|image        = [[Image:Circularity1.jpg|200px]]
|header1 = Iris Segmentation Code
|header2 = '''Contact:''' [[Fernando Alonso-Fernandez]]
}}


=='''Introduction'''==

'''This page provides a software code for iris segmentation based on the Generalized Structure Tensor (GST)''', based on publications [1] and [2] ([http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References below] in this page). Its development was made mainly during the 2011-2013 [http://islab.hh.se/mediawiki/BIO-DISTANCE FP7 Marie Curie IEF Action BIO-DISTANCE], but it contains some improvements and add-ons incorporated afterwards at different points in time. We keep working constantly in this topic, so if you are interested, it is maybe worth checking from time to time if new improvements have been added to the code.

The software accepts an iris image as input, and outputs segmentation information of the input iris image (see below for more information). It is capable of handling images acquired both in near-infrared (NIR) and visible (VW) spectrum.


'''The GST code consist of the following steps''' (some can be deactivated or customized, please read the documentation included with the code):

1) <u>Image downsampling</u> for speed purposes. This will not jeopardize accuracy, since the detected iris circles are later fitted to the irregular iris contours, so any loss of resolution in iris circles detection due to downsampling is compensated.
<!-- Esto es un comentario -->

2) <u>Contrast normalization</u> based on imadjust (Matlab function). This increases the image contrast, spreading grey values fully in the 0-255 range.
<!-- Esto es un comentario -->

3) <u>Specular reflection removal</u> based on the method published in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 3].
<!-- Esto es un comentario -->

4) <u>Computation of the image frequency</u> based on the method published in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 2]. This helps to customize inner parameters of steps 5-8 to the input image.
<!-- Esto es un comentario -->

5) <u>Adaptive eyelash removal</u> using the image frequency, as indicated in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 2]. The method is based on p-rank filters as published in reference [4]. Eyelashes are removed since they create strong vertical edges that may mislead the filters used for eye center estimation and iris segmentation in steps 7 and 9.
<!-- Esto es un comentario -->

<br />
[[Image:gst_iris_image_prepro.png|thumb|none|center|middle|600px|Image preprocessing]]
<br />

6) <u>Adaptive edge map computation</u> using the image frequency, as indicated in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 2]. Edge map is the basis for eye center estimation and iris boundaries detection, see references [1, 2] for further details.
<!-- Esto es un comentario -->

7) <u>Estimation of the eye center</u> based on the method published in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 2] using circular symmetry filters. The estimated center is used to mask candidate regions for the centers of iris circles, helping to improve detection accuracy in step 9.
<!-- Esto es un comentario -->

8) <u>Detection of eyelids</u> based on linear symmetry detection of horizontal edges (unpublished and unoptimized, only return a straight line). The detected eyelids are used to mask candidate regions for the centers of iris circles too, helping to improve detection accuracy in step 9.
<!-- Esto es un comentario -->

9) <u>Detection of iris boundaries</u> based on the method published in [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 1] using the Generalized Structure Tensor (GST). In NIR images, the inner (pupil) circle is detected first, while in VW images, the outer (sclera) filter is detected first. This is because in NIR images, pupil-to-iris transition is sharper than iris-to-sclera transition, thus more reliable to detect in the first place. The opposite happens with VW images.
<!-- Esto es un comentario -->

10) <u>Irregular contour fitting</u> based on active contours as published in reference [http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References reference 5].

<br />
[[Image:gst_iris_image_segment.png|thumb|none|center|middle|600px|Image segmentation]]
<br />

'''The code outputs the following information of the input iris image''':
* Segmentation circles of the iris region (inner and outer circle) as well as eyelids (in the form of a straight line)
* Irregular (non-circular) iris boundaries fitted by active contours 
* Estimated eye center (computed at the beginning and used to assist in the segmentation)
* Intermediate images after contrast normalization, specular reflection removal, and eyelash removal
* Complex edge map of the input image
* Binary segmentation mask


=='''Terms and Conditions'''==

This code has not any warranty and it is provided for research purposes only. 

The code is provided in the form of executables compiled with Matlab r2009b 32 bits (mcc command) under Windows 8.1. It accepts as input grayscale and RGB images in any format supported by Matlab "imread" (uint8 only).

Certain parameters of the code are customizable, please read the documentation included with the code for more information.

'''Please remember to cite reference [1] ([http://islab.hh.se/mediawiki/index.php/Iris_Segmentation_Code_Based_on_the_GST#References below]) if you make use of this code in any publication'''.

By downloading the code, you agree with the terms and conditions indicated above.

'''Download the code''' [http://islab.hh.se/mediawiki/index.php/Fernando_Alonso-Fernandez here] (latest release: July 2015)


=='''People responsible'''==

* [http://islab.hh.se/mediawiki/index.php/Fernando_Alonso-Fernandez Fernando Alonso-Fernandez] (contact person)
* [http://islab.hh.se/mediawiki/index.php/Josef_Bigun Josef Bigun]


=='''References'''==

# F. Alonso-Fernandez, J. Bigun, “Iris Boundaries Segmentation Using the Generalized Structure Tensor. A Study on the Effects on Image Degradation”, Proc. [https://sites.google.com/a/nd.edu/btas_2012/ Intl Conf on Biometrics: Theory, Apps and Systems, BTAS,] Washington DC, September 23-26, 2012 ([http://hh.diva-portal.org/smash/record.jsf?searchId=2&pid=diva2:545745 link to the publication])
# F. Alonso-Fernandez, J. Bigun, “Near-infrared and visible-light periocular recognition with Gabor features using frequency-adaptive automatic eye detection”, [http://digital-library.theiet.org/content/journals/iet-bmt, IET Biometrics], Volume 4, Issue 2, pp. 74-89, June 2015 ([http://digital-library.theiet.org/content/journals/10.1049/iet-bmt.2014.0038 link to the publication in IET Biometrics])
# C. Rathgeb, A. Uhl, P. Wild, "Iris Biometrics. From Segmentation to Template Security", Springer, 2013
# Z. He, T. Tan, Z. Sun, X. Qiu, "Toward accurate and fast iris segmentation for iris biometrics", IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 31, (9), pp. 1295–1307 ([http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4586378&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4586378 link to the publication in IEEE Xplore])
#J. Daugman, "New methods in iris recognition", IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, 37(5), 2007 ([http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4305270&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4305270 link to the publication in IEEE Xplore]) ([http://www.cl.cam.ac.uk/~jgd1000/NewMethodsInIrisRecog.pdf link to the publication in the author´s web site])