Development of Sclera Biometric system using Machine learning approach

Abstract

The biometric systems are becoming more popular in the present days because of the newlinesecurity requirement in the field of data, e-commerce, business, military, internet and electronic transfers. There are many biometric systems available to authenticate an newlineindividual like face, fingerprint, iris recognition etc, but there are some limitations in newlinethese systems. To overcome such limitations, this research focuses on sclera newlinerecognition, the most accurate form of biometric identification. The robustness of newlinesclera recognition comes from the individual characteristics of the human, and the newlinepermanency of the sclera texture, which is stable over human life, and the newlineenvironmental effects cannot easily change its shape. Identification of a person based newlineon the obtained unique set of feature is a significant task. The human identification is newlinefeasible with many biometric systems and the sclera recognition is the more accurate and best biometrics. The sclera is the white portion of an eye . The vein pattern present in the sclera portion is distinct in every individual. Hence, the sclera vein pattern is a suitable biometric technology used for the human identification process. Only frontal looking images are considered in the existing methods used for sclera newlinerecognition which is a major limitation. The sclera recognition is proven to be a newlinepromising scheme but it is slow in matching process, so to speed up the matching process Neural Network (NN) method is adopted to segmentation of the images. This research work presents a concept called sclera recognition, which focuses on: preprocessing technique, feature extraction and then classification technique like neural network for sclera biometric. Therefore, the goal of this research work is to design an accurate sclera recognition scheme. In this regard, this work has two contributions like Multi-angle Sclera biometric scheme and DWT-HJDLBP scheme , which are aimed at resolving computation issues of the sclera recognition. newline