Biometrics is a technique that analyses and measure human body characteristics such as voice patterns, DNA, eye retinas, facial patterns, irises for authentication. A palm vein authentication system uses blood vessel patterns as key for personal identification. This paper intends to design architecture by palm vein analysis technique. Biometric analysis can be done with suitable looping algorithms like masking and mapping, a novel technique of extracting vein features out of palm which posses unique identity for different users. An IR-Illuminator is used to illuminate palm veins and is captured using image sensor. The raw image data so obtained should be pre-processed which includes segmentation of veins with respect to muscles, removal of noise, alignment of palm and orientation of veins. The pre-processing is simulated on MATLAB. A 2-level secure architecture for encoding and decoding process in Xilinx has been developed and can be dumped to FPGA also. The architecture takes bit data as input from MATLAB and performs verification of palm vein Region Of Interest (ROI) at individual levels and sends back the decoded bit to MATLAB for display on monitor for user. This project is robust in the sense image data acts as first level security and user input password converted to CRC-16 CCITT constructs a polynomial there by encoding image data this way a 2-level security and good authentication to the user is provided. In decoding process, the encoded polynomial is divided by image data set and if the reminder is zero the authentication to user is provided else a error message is displayed.
