The proposed method is one of the possible methods of personal identification based on an iris image recognition. The method is based on evaluating the similarity of two objects: the iris of the identified person and the reference iris pattern.

Nowadays there is a huge requirement for protection systems (access control systems, data protection systems, etc.). This caused a necessity of elaborating methods and algorithms of personal authorisation.
The most common approaches to the access control are methods using natural, individual features of each person (e.g. fingerprint identification, voice recognition, eye scanning, signature verification). All these methods are based on the comparison between a feature and a reference pattern. The reference patterns are the previously registered data about features of people authorised to access the system. Data of each person willing to enter the system must be compared with all patterns in the database for identification. The picture and the pattern may be translated, rotated and scaled, it is necessary to find unchanging features of this objects - the invariants of these geometric transformations.
In this paper we present one of the possible methods of personal identification based on the iris image recognition. The method is based on defining similarity between the iris of identified person and the reference one (we used the Fourier-Mellin transform). The uniqueness of the iris pattern allows to identify a person.
Based on this approach personality and medical diagnostic systems may be realised.