New M.E. Thesis Submitted from CSE Student


Biometric recognition is a reliable way to authenticate the identity of a living person based on physiological or behavioral characteristics. Iris recognition systems make use of the uniqueness of the iris patterns to derive a unique mapping where it becomes possible to apply some matching algorithms to identify a person. Iris recognition, as a biometric method, outperforms others because of its high accuracy. Iris recognition has the ability to handle very large populations at high speed. However, this technology still suffers from problems. One problem is to independently evaluate the algorithm and to automate the recognition of the iris by reducing complexity and increasing algorithm speed. Numbers of other challenges are also faced while working with the iris recognition system. In iris recognition, a three-stage approach is widely used by researchers. These stages are preprocessing, feature extraction and recognition stage.
In the present work, many methods are combined to build a reliable and fast method for feature extraction in iris recognition system. Reliable techniques for iris image enhancement and circle detection are used. These techniques can then be used to facilitate the further study of the statistics of iris. Also a program coding with MATLAB going through all the stages of the iris recognition is built. It is helpful to understand the procedures of iris recognition and demonstrate the key issues of iris recognition. The Hamming distance has been employed for classification of iris templates, and two templates have been found to match if a test of statistical independence failed. The system performed with perfect recognition and resulted in false accepts and false reject rates of 0.01% and 0.61% respectively. The accuracy of the system is found to be 99.38%. Therefore, iris recognition is reliable and accurate biometric technology.

Leisure Readings :