Hand identification using fuzzy feature matching (FFM)

Abstract

The rapid growth in the use of ecommerce applications requires security of the communication channel used and reliable and auto personal identification for effective security control. Conventional identification techniques can be divided into two categories: token based, such as physical key, an ID card and a passport, and knowledge-based, such as a password in this paper we propose a WAVELET based hand identification system for personal identification. These approaches offer more scope for credible authentication as compared to finger print recognition. The images of the hand are taken and length of the fingers is computed using gradient method. These are used as feature for categorizing the palm. The variance of the lines in the palm is calculated for all sub blocks of the palm image to get another set of feature vectors. Therefore the numbers of parameters, which have to be matched for authentication, are more in the case of palm than a finger. From these extracted features we take the minutiae of the lines in the palm. This paper proposes a novel method, a fuzzy feature match (FFM) based on a local triangle feature set to match the deformed palm. The palm is represented by fuzzy feature set: the local triangle feature set. The similarity between the fuzzy features set is used to characterize the similarity between the palms. A fuzzy similarity measure for two triangles is introduced and extends to construct a similarity vector including the triangle-level similarities for all triangles in two palms. Accordingly, a similarity vector pair is defined to illustrate the similarities between two palms. The FFM method maps the similarity vector pair to normilized value, which quantifies the overall image similarity. (7 pages)