Extraction of Unique Feature Sets (Bio-metric Identifier) For Human Ear’s Image using DWT Analysis

The presented paper presents the statistical analysis of the dwt coefficients of ear’s images in order to ascertaining the bio-metrics identity of a person. The dwt coefficients matrix contains the frequency variations of ear’s images in three parts; the lower half of the matrix that contains the low frequency elements that make up the flashy part of the ear, the upper half that contain the high frequency elements in form of noise and diagonal elements that contain the maximum information about the features of the ear’s image. The diagonal elements or dwt coefficients are then analyzed statistically in terms of standard deviation, entropy and covariance to identify the given ear’s image.