Iris Recognition for Biometric Identification using Dyadic Wavelet Transform Zero-Crossing D. de Martin- Roche*, C. Sanchez-Avilat & R. Sanchez-Reillot * Ericsson Espaiia S.A., Dpt. Software Engineering, C/ Ombu, 3, 28045 Madrid, Spain. ‘Dpto. de Matematica Aplicada, E.T.S.I. Telecomunicacion,Universidad Politecnica de Madrid, 28040 Madrid, Spain. Dpto. de Ingenieria Electrica, Electronica y Autornatica, Universidad Carlos I11 de Madrid, 28911 Madrid, Spain. Abstract - In this work, a new biometric identifica- tion approach based on the human iris pattern is pro- posed. The main idea of this technique is to represent the features of the iris by fine-to-coarse approximations at different resolution levels based on the discrete dyadic wavelet transform zero-crossing representation. The re- sulting one-dimensional (1-D) signals are compared with model features using different distances. Before perform- ing the feature extraction, a pre-processing step is to be made by image processing techniques, isolating the iris and enhancing the area of study. The proposed tech- nique is translation, rotation and scale invariant. Results will show a classification success above 98% achieving an Equal Error Rate equal to 0,21% and the possibility of having null False Acceptance Rates with low False Re- jection Rates. Introduction Nowadays, one of the main threats that IT systems and security environments can have, is the possibility of hav- ing intruders in the system. This is normally solved by user authentication schemes based on passwords, secret codes and/or identification cards or tokens. Schemes based only on passwords or secret codes can be cracked by intercepting the presentation of such a password, or even by counterfeiting it (via passwords dictionaries or, in some systems, via brutal force attacks). On the other hand, an intruder can attack systems based on identifi- cation card or tokens by robbing, copying or simulating them. If the scheme used in the system is based both on a card and a password (usually called Personal Identifi- cation Number - PIN), the intruder should apply more effort to gain entry to the system, and with more ad- vanced technologies, such as smart cards, some vulnera- bilities of the system could be avoided (e.g. brutal force attacks are impossible under a well-defined smart card). As it is well-known, biometrics deals with identifica- tion of individuals based on their biological and/or be- havioral features. Technologies that exploit biometrics have the. potential application of identifying individuals in order to control access to secured areas or materials. Nowadays a lot of biometric techniques are being de- veloped based on different features and algorithms. In fact, there are many biometric techniques that are ei- ther widely used or under investigation, including voice, face, iris, fingerprint, ear, retinal scan, signature, etc. [3]. Each technique has its strengths and limitations, not being possible to determine which is the best. No sin- gle biometrics is expected to effectively meet the needs of all the applications. Nevertheless, it is known that, from all of these techniques, iris recognition is the most promising for high security environments (21. The bio- metric identification problem can be categorized into two fundamentally distinct types of problems with different complexities: recognition (or identification) and verifica- tion (or authentication). Recognition refers to the prob- lem of establishing a user’s identity. Verification refers to the problem of confirming or denying a user’s claimed identity. The possibility that the human iris might be used as a kind of optical fingerprint for personal iden- tification was suggested originally by ophthalmologists. Therefore, the potential of the human iris for such kind of problems comes from the anatomy of the eye. Some properties of the human iris that enhance its suitability for use in automatic identification include: 1) its inherent isolation and protection from the external environment, being an internal organ of the eye, behind the cornea and the aqueous humor; 2) the impossibility of surgically modifying it without high risk of damaging the user’s vi- sion; and 3) its physiological response to light, which provides the detection of a dead or plastic iris, avoiding this kind of counterfeit. Also several studies have shown that while the general structure of the iris is genetically determined, the particulars of its minutiae are critically dependent on initial conditions in the embryonic meso- derm from wich it develops. Therefore, there are not ever two irises alike, not even for uniovular (identical) twins [2]. In these respects the uniqueness of every iris paral- lels the uniqueness of every fingerprint. At the moment, only two prototype iris-recognition systems had been de- veloped by Daugman [a] and Wildes et a1.[9]. Neverthe- less, this biometric technology presents still many open problems. So, more recently, we can find some works in this issue, as given by Boles et al.[l], using wavelet transform, and by Sanchez-Rei110 et al. [7], where Gabor filters are used. Here, we develop a new approach using dyadic wavelet transform zero-crossing. A wavelet func- tion that is the first derivative of a cubic spline will be used to construct the representation. Image Acquisition and Preprocessing 0-7803-6636-0/01/$10.00 02001 IEEE 272