An Evaluation of Palm Print Recognition Techniques using DCT, Haar Transform and DCT Wavelets and their Performance with Fractional Coefficients
Aditya A. Tirodkar, Tanuja Sarode, Hemant B. Kekre · 2011
print recognition is a proliferating technique in the ever increasing use of Biometrics for security and identification. Our paper proposes a technique for palm print recognition using the Discrete Cosine Transform, Haar Transform and the DCT Wavelet along with tests related to their fractional coefficients. After experimental observation on a set of 8000 images, we have found their maximum accuracies to be in the range of 94.4% for each transform. Along with discussing each transform used, we demonstrate how partial energy coefficients can increase accuracy with a 99.5% reduction in computation, as can be inferred from our results.