An efficacious matching of finger knuckle print images using gabor feature

Abstract

An efficacious matching of 240 Finger knuckle print images using Gabor feature by extracting position and weight with a record of 5362 values for each image. In this approach, the original image is compared with the other images captured at different angles. According to the obtained records, a difference is calculated between the two images which results in how much disparity subsist between the two images. The comparison is made apparent by generating a 3-D graph which shows the best match of the original image. The data is taken from Hong Kong Polytechnic University which contains 7920 FKP images of 660 different individuals.