Iris recognition using the JAVAVis Library

This project has been created to develop a biometric identification system through a man's iris using a computer to perform the processing of the pictures. To develop this application, and to differentiate the project from others who have already implemented, we have used the image processing library JAVAVis and JAVA as a programming language. I.INTRODUCTION Biometrics is based on the premise that each individual is unique and has distinctive physical traits or behaviours, which can be used to identify or validate him/her. The Iris human structure is only one per person and it forms a very complex system unchanged during any person's life. For several years the human iris has been using for user authentication. To this, an image capture is necessary. It must be transform into black-white ones, in an environment properly lighted, using a high-resolution camera. Usually this is done by looking through the a fixed camera lens, a person simply positioned in front of the camera and the system captures the iris image, this image is processed to extract Biometry optical iris patterns, which in turn are subject to mathematical algorithms to produce a sufficient amount of data to identify ambiguously of the individual concerned. That is compared with another sample taken earlier and stored in the database system, so that if both match the user is considered authenticated successfully, the probability of a false acceptance is the lowest of all biometric models. We start from the idea that in a future the application could be used by capturing images through a webcam, but now and ought to technical limitations, images prosecuted are .JPG files previously entered into the computer, which must have a high resolution to achieve a smooth running of the program. The image, of which we are referring to, is digitally processed to obtain a code that characterized it. They are used in the treatment pre-processing algorithms to enhance image quality and thus, extract higher quality information and therefore more accurate. The code that characterizes an iris image is obtained by applying Haar wavelet transform type. The iris code is composed of 366 bits and thus obtained 1.5031x10110 different codes. Evidence of encryption algorithms have been conducted on internet images. 1. Capture the image. 2. Pre-processing image with the help of JAVAVis library. 3. Removing the area of interest. 4. Haar wavelet application. 5. Compare with codes from the database. At the beginning of …