u ser guide - nist
TRANSCRIPT
Nation
Info
U
nal InstituInformat
ormation
User
M
YooyRoss
P. Jona
ute of Station TechAccess D
r Gu
ay 2011
young LeJ. Micheaathon Phi
ndards anology La
Division (IA
ide
e als llips
nd TechnaboratoryAD) / Ima
nology (Ny (ITL) age Group
IST)
p
1
Disclaimer for Video‐based Automated System for Iris Recognition (VASIR)
May 23, 2011
The software was developed by the National Institute of Standards and Technology (NIST), an agency of the Federal Government. Pursuant to Title 15 United States Code Section 105, works of NIST are not subject to copyright protection in the United States and are considered to be in the public domain. Thus, the software may be freely reproduced and used. Please explicitly acknowledge the National Institute of Standards and Technology as the source of the software.
This software is released by NIST as a service and is expressly provided "AS IS." NIST MAKES NO WARRANTY OF ANY KIND, EXPRESS, IMPLIED OR STATUTORY, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTY OF MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT AND DATA ACCURACY. NIST DOES NOT REPRESENT OR WARRANT THAT THE OPERATION OF THE SOFTWARE WILL BE UNINTERRUPTED OR ERROR‐FREE, OR THAT ANY DEFECTS WILL BE CORRECTED.
Certain trade names and company products are mentioned in the text or identified. In no case does such identification imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the products are necessarily the best available for the purpose.
With the exception of material marked as copyrighted, information presented in this document is considered public information and may be distributed or copied. Use of appropriate by line/photo/image credits is requested.
2
Contents
Purpose ............................................................................................................................... 3
Requirements ...................................................................................................................... 4
Configuration .................................................................................................................. 4
Prerequisites ................................................................................................................... 4
Installation ...................................................................................................................... 4
Getting Started .................................................................................................................... 8
Cascade menu ................................................................................................................. 8
Detect menu ................................................................................................................... 8
Extracting the eye region image ................................................................................... 10
Match tab ...................................................................................................................... 11
Quality menu ................................................................................................................. 14
Match menu .................................................................................................................. 14
Annex A. Doxygen description of source code ................................................................. 16
Annex B. MBGC (Multiple Biometric Grand Challenges) .................................................. 16
3
Purpose
This document gives an overview on how to operate VASIR. The documented source‐code and information regarding the MBGC dataset can be found in the appendix.
The tool can be used to:
▪ Detect the eye pair (left and right eyes) within their video data
▪ Provide the left and right ocular region for ocular image quality analysis
▪ Identify a person using the iris feature
▪ Study still‐ and video‐based iris recognition
The system is under development and optimization. Therefore, it is to be understood that the source code produces warnings and is not without bugs. We plan to update each component of VASIR after evaluating its quality and performance.
Please note that we cannot warrant the correctness, usefulness, accuracy, reliability, etc. of the source code. We would, however, appreciate if you could send BUG REPORTS to vasir@<NOSPAM>nist.gov.
4
Requirements
Configuration
The tool has been tested under following operating systems:
▪ Microsoft Windows® XP Professional Edition
▪ Microsoft Windows Server 2008
▪ Microsoft Windows 7 (x86)
The recommended minimum configurations are as follows:
▪ A dual core CPU: 2GHz
▪ RAM: 3GB
▪ Display resolution: 1280 x 1024 pixels
Prerequisites
Qt UI framework
Download the latest version at: http://qt.nokia.com/downloads
(e.g., Qt libraries 4.7.3 (exe, ~230MB) would take about 2 hours installation time depending on your system’s performance)
OpenCV Library
Download version 1.0 from http://sourceforge.net/projects/opencvlibrary/files/opencv‐win/
(e.g., OpenCV_1.0 (exe, ~18MB) would take about 1‐5 minutes installation time)
For the detailed information, see http://opencv.willowgarage.com/wiki/InstallGuide
Visual Studio Pro 2005 or 2008
After the installation, please update the Microsoft Visual C++ 2005 (or 2008) Redistributable Package (x86).
The user may be required to install the VsQtAddIn_2005 (or 2008).msi to integrate Qt’s UI tool into Visual Studio. Please refer to the addin’s website at http://mm‐werkstatt.informatik.uni‐augsburg.de/Volker‐Wiendl.html
More recent version of Qt supports the Visual Studio addin out‐of‐the‐box; we did not test the updated Qt addin with this source code. Please have a look at http://qt.nokia.com/.
VASIR source code (Written in C++)
Download the source code “NIST_VASIR_src_beta_v1.0 (ZIP, ~200KB)” at: http://www.nist.gov/itl/iad/ig/vasir.cfm
* NOTE: You may choose to use a different operating environment (e.g. Mac and Linux). However, the installation steps might vary depending on your configuration.
5
Installation
The following steps exemplify the installation with Visual Studio Pro 2005 on Windows XP for the developer:
Qt, OpenCV, and VS2005 on Windows XP 1. Create the folder “C:\Qt” 2. Extract “qt‐win‐opensource‐src‐version.zip” into “C:\Qt\” 3. Add QTDIR “..\Qt\bin\” to the “Path” Environment variable: System Settings > System > Advanced > Environment variables 4. Open the “Visual Studio Command prompt” 5. Change the path to “C:\Qt\ 6. Run “configure –platform win32‐msvc2005” 7. Run “nmake” 8. Create the folder “C:\OpenCV\” 9. Run OpenCV_1.0.exe into “C:\OpenCV\” 10. Start Visual Studio 2005 11. Check the path of your Visual Studio version, click “Solution ‘YooIRIS’”, and open “Properties > Common Properties > Debug Source Files” 12. Open “Tools > Options > Projects and Solutions > VC++ Directories” 12.1 Add the following components into “Include directories”: C:\OpenCV\cv\include C:\OpenCV\cxcore\include C:\OpenCV\cvaux\include C:\OpenCV\otherlibs\highgui C:\Qt\include 12.2. Add the following components into "Library files": C:\OpenCV\lib C:\Qt\lib 12.3. Add the following components into "Source files": C:\OpenCV\cv\src\ C:\OpenCV\cxcore\src C:\OpenCV\cvaux\src C:\OpenCV\otherlibs\highgui 13. Click the “YooIRIS” project and change the Qt version from “Properies > C/C++ and Linker” 14. Close Visual Studio 2005 15. Install “VsQtAddIn_2005.msi” to "..\Visual Studio 2005\Addins\" e.g., C:\Documents and Settings\User Name\My Documents\Visual Studio 2005\
6
The following steps exemplify the installation with Visual Studio Pro 2008 on Windows 7 for the developer:
Qt, OpenCV, and VS2008 on Windows 7 1. Create the folder “C:\Qt” 2. Extract “qt‐win‐opensource‐src‐version.zip” into “C:\Qt\” 3. Add QTDIR “..\Qt\bin\” to the “Path” Environment variable: System Settings > System > Advanced > Environment variables 4. Open the “Visual Studio Command prompt” 5. Change the path to “C:\Qt\ 6. Run “configure –platform win32‐msvc2008” 7. Run “nmake” 8. Create the folder “C:\OpenCV\” 9. Run OpenCV_1.0.exe into “C:\OpenCV\” 10. Start Visual Studio 2008 11. Check the path of your Visual Studio version, click “Solution ‘YooIRIS’”, and open “Properties > Common Properties > Debug Source Files” e.g., change “Microsoft Visual Studio 8” to “Microsoft Visual Studio 9” 12. Open “Tools > Options > Projects and Solutions > VC++ Directories” 12.1 Add the following components into “Include directories”: C:\OpenCV\cv\include C:\OpenCV\cxcore\include C:\OpenCV\cvaux\include C:\OpenCV\otherlibs\highgui C:\Qt\include 12.2. Add the following components into "Library files": C:\OpenCV\lib C:\Qt\lib 12.3. Add the following components into "Source files": C:\OpenCV\cv\src\ C:\OpenCV\cxcore\src C:\OpenCV\cvaux\src C:\OpenCV\otherlibs\highgui 13. Click the “YooIRIS” project and change the Qt version from “Properies > C/C++ and Linker” 14. Close Visual Studio 2008 15. Install “VsQtAddIn_2008.msi” to "..\Visual Studio 2008\Addins\" e.g., C:\Documents and Settings\User Name\My Documents\Visual Studio 2008\
7
VASIR Source code 1. Unzip “NIST_VASIR_src_beta_v1.0.zip” and open the source code 2. Select “YooIRIS.ui” and”ModeDialog.ui” and open its context menu (right click) 3. Adjust the path “Command Line” in “Properties > “Custom Build Step” to “C:\Qt/bin/uic.exe” 4. Repeat steps 2. and 3. for “ImageWidget.h”, “ModeDialog.h”, and “YooIRIS.h” files 5. Adjust the path “Command Line” in “Properties > “Custom Build Step” to “C:\Qt/bin/moc.exe” 6. Build the source code *Warning: this source code is not complete version. Note that you may get a known warning or memory related messages.
Gett
Casca
The “Cdone “Matc
A clas
Click P
Note
Dete
The ““Open
Depe
Click “
ting Start
ade menu
Cascade” mebefore loadinch tab”).
ssifier named
Pick “Select c
that dependi
ct menu
Detect” menun CAM” are c
nding on the
“Open FACE A
ted
nu allows theng a video file
“parojosG.xm
ascade” and
ng on your go
u allows the uurrently disa
dataset, the
AVI” and load
e user to loade. (If you wan
ml” can be fo
load the “par
oal a differen
user to load abled.
installation o
d the face visi
8
d a Haar‐classt to load a sti
und in “/bin/
rojosG.xml” c
nt, custom cla
a video file in
f additional c
ble video file
ifier (XML forill image dire
/cascade/” wi
classifier.
assifier can be
AVI format.
codecs may b
to start the e
rmat). This nectly, please se
thin the defa
e used.
Both “Open I
e necessary.
eye regions d
eeds to be ee Section
ault folder.
IRIS AVI” and
detection.
Once the ey
Relev
1
2
3
the detectionye region det
ant messages
1) Number of
2) Image quali
3) Failed detec
n is started, tecting proces
s are displaye
total frames
ity scores for
ction
he “Detectionss on Detectio
ed in the com
in the selecte
the left and r
9
n” tab will dison tab ‐ as sh
mand (= cons
ed video file
right iris imag
splay all videohown in the fig
sole) window
ges
o frames alongure below.
w below;
ngside with
After right etab.
The “a vide
Extra
The syfolder
“<Vid
a successful reye based on
Data Type heeo file. See de
acting the ey
ystem automr as the origin
eoFileName>
run, the systen the calculate
ere” was chanetails in Sectio
ye region im
atically extranal video file.
>_<Frame #>_
em automaticed quality sco
nged to “Distaon “Match ta
mage
cts the left anThe naming f
_<L(left)|R(rig
10
cally selects thores and displ
ant Video Frab”.
nd right eye iformat is as f
ght)><Sequen
he best qualitlays the selec
me” since the
mages and safollows:
nce #>.bmp”
ty images forcted images in
e imagery wa
aves them to
r the left and n the “Match
as loaded from
the same
”
m
Matc
The “
If youcasca
VASIR
▪
▪
▪
▪
A quic
A
B
C
D
ch tab
Match” tab is
would like tode and the AV
R allows multi
Distant Vid
Distant Vid
Classical St
Classical St
ck glance of t
. Click this bbe auto‐sc
. Will displa
. Click this bsample.
. Will displa
A
B
s used for iris
o match a stilVI video file f
iple scenarios
deo to Distan
deo to Classic
till to Classica
till to Distant
the screensho
button – or ancaled and disp
y the data typ
button to mat
y the matchin
C
verification (
l image only, firsthand.
s to verify two
t Video
cal Still
al Still
Video
ot below:
ny other “Opeplayed within
pe that you s
tch the image
ng results for
11
(1:1).
you can use
o biometric s
en …” button the box belo
elected in the
es between th
r the left and
this tab witho
amples depe
‐ to load an iow the button
e “Select Mod
he target and
right iris imag
D
out having to
nding on you
iris image. Thn.
de” dialog.
the query bi
ges.
o load the
ur purpose:
he image will
ometric
If youmode
The “Ctaken
The “on Mo
Pick t
click the “Ope” that contai
Classical Still by a system
Distant Videoove (IOM) sys
he type and t
pen Left (or Rns two radio
Image” radioakin to an LG
o Frame” indistem at a cert
then click “OK
Right) Image”buttons: “Cla
o button indicG2200 or LG40
cates the iris tain distance.
K.”
12
button, you wassical Still Im
cates that the000.
image is actu.
will be shownmage” and “Di
e iris image is
ually a video f
n a dialog titleistant Video f
of decent qu
frame, captur
ed “Select frames”.
ality. It was
red by the Iriss
At thi
Now,
The scscenathe rig
s point load t
click “Go!” to
creenshot shoario. The two ght eye are fr
the iris image
o get the mat
ows an examiris images forom a differe
e; supported f
tching results
ple of the Disor the left eyent person.
13
formats are B
.
stant Video Fre are from the
BMP, TIFF, JPG
rame vs Classe same perso
G, and PGM.
sical Still Imagon while the tw
ge matching wo images foor
Quali
Click tqualit
Matc
1
2
3
ity menu
the “Choose”ty score in the
ch menu
. Click “Cho
. Select the
. Load the ta
” menu item te command w
ose” to matc
mode for the
arget image.
to assess the window.
h a target stil
e target image
14
quality of an
l image and a
e.
image. You w
a query still im
will see the ca
mage.
alculated
4
5
You w
Ack
We w
. Select the
. Load the q
will see the m
nowledg
would like to th
mode for the
query image.
atching result
ement
hank Kevin C
e query image
ts in the “Ma
C. Mangold fo
15
e.
tch” tab.
or his valuable
e comments oon this user guuide.
16
Annex A. Doxygen description of source code
Please have a look “VASIR_doxygen_html” file on our website: http://www.nist.gov/itl/iad/ig/vasir.cfm
Annex B. MBGC (Multiple Biometric Grand Challenges)
The VASIR components were evaluated using the datasets collected by the Multiple Biometric Grand Challenge (MBGC). Two goals of the MBGC are to promote and to advance iris recognition technology. The MBGC dataset consists of eye images of varying illumination conditions, low quality, and off‐angle or occluded images in both still and video imagery.
For further information and a downloadable version of the dataset, please refer to the website: http://face.nist.gov/mbgc/