Page 2 - Iris Recognition. Need An Iris Database, Please

the first stage in this project is Iris localization in which you need to automatically extract the Iris texture. this stage can be provided using more than one algorithm:
- Hough transform (used by Libor Masek).
- Dr.daugmans' algorithm (I used this ).

if you decide to use Hough transform you must preprocess the iris image using canny edge detection and nonmaxima suppression.
Hough transform is simply finds the circles in any image but the input for this transform must be binary image thats why we used the canny edge detection and nonmaxima suppresion.
(by the way there is a good book for all this concepts : feature extraction and image processing by Mark S. Nixon. Alberto S. Aguado ).

john daugmans' algorithm just look for the circle that have maximum deference.

imagine that you have a black image that have a black circle enclosed by white circumstance with r radius where
rmin < r < rmax when you look for all black circles (from rmin to rmax) that have maximum defernce (the deference is between the circle with r1 and the circle with r1+1)you will find just one circle which enclosed by the white line this is the idea (if you do not understand I will explain it later)

Thankyou very much Eyaad for your reply. Dude, shall i add you on google talk, so tat we can chat live. i need some help with the simulation of Libor Masek's code. i get a plethora of errors when i run it. It would be gr8, if u could help me with the simulation of this program. i have added u..here's ma gmail id: arun.nair87@gmail dot com

did anyone have CASIA iris database?
please forward it to me because i have some difficulties using my own image that i caputre my self..
thx
my email : handy_venom01@yahoo dot com

hi all
really iam new here at this great forum , my name is mohammad mustafa studying @ Birzeit University in Palestine Electrical Engineer/Graduation year and my Graduation project is about Image processing ..
in our step now we are to detect the iris color ,firstly must doing iris detection .. anyway i used Duaghman technique its working well , now the problem how to find the color of the iris .. my idea is to find the average value of the pixles exists between the iris circle and the pupil circle ( iam usong RGB images ),, plz if u can help me about how to extract the interested area and how to find its avg pixels value ) because iam tired about searching and working in matlab ,,
in the other hande ,, iam good at the image processing work if any one need somethings about human body detection features iam ready to help ( fingerprint , face , skin .. ) ...
with respect
Mohammad Mustafa
Elect.Eng
m.saifeddinathotmaildotcom ( 15 hours online )
atyahoodotcom
atgmaildotcom

hi there its ok i didnt see any reply .. but to simplyify my need ..
in daugman matlab code ,, how can i after finding the two circles how can i convert the desired shape to rectangual !
plz i need urs expirence to help me
bye

use normalization equation given by J.Duagman its simple.
its like to pick number of points radialy (from pupil circumstance to iris cirumstance )with zero angle and align them up to down in first column of the matrix of the output rectangle. repeate this with another angle (say two degrees) and fill the second column, finally you find your self convert the radial dimensions to rectangular dimensions.
كيف حالك ياطيب

اياد حبيبك ضيفني بدي احكي معك ازا بتريد لاني ما فهمت عليك كويس رغم اني رح احاول افهم بزيادة شو كتبتلي لاتنسى تضيفني بستناك وحياتك m.saifeddin@hotmail.com

hi everone,

i doing a project on fake iris detection system, so i need a fake iris database or some one having iris camera ....plz help me
mail me at
aishwarya_titan at yahoo dot co dot in

السلام عليكم

أخي اياد انا بحاجة الى مساعدتك ..

انا بصدد البدء في مشروع تخرجي وهو حول " بصمة العين "

ولدي الكثير من الاستفسارات ..

فارجو ان تساعدني

technician_1 at hotmail dot com

ادهم /غزة

hi aiswaria
i too doing the project....
pls help me to correct the code ....
hogh circle doesn't work .....
any simple programs for that
pls reply me in aneeshrp at yahoo dot com