Personal Identification Using Palmprint and Contourlet Transform

  • Authors:
  • Atif Bin Mansoor;M. Mumtaz;H. Masood;M. Asif Butt;Shoab A. Khan

  • Affiliations:
  • National University of Science and Technology, Rawalpindi, Pakistan;National University of Science and Technology, Rawalpindi, Pakistan;National University of Science and Technology, Rawalpindi, Pakistan;National University of Science and Technology, Rawalpindi, Pakistan;National University of Science and Technology, Rawalpindi, Pakistan

  • Venue:
  • ISVC '08 Proceedings of the 4th International Symposium on Advances in Visual Computing, Part II
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Palmprint based personal verification has gained preference overother biometric modalities due to its ease of acquisition, highuser acceptance and reliability. This paper presents a newpalmprint based identification approach which uses the texturalinformation available on the palmprint by employing the ContourletTransform (CT). Center of the palm is calculated using the DistanceTransform and by calculating the parameters for the best fittingellipse, the alignment of hand `θ ' is found.Rotational invariance is achieved by cropping a square region ofsize 256 x 256 pixels around the center aligned at θ degrees. After establishing the region of interest (ROI), the twodimensional (2-D) spectrums is divided into fine slices, usingiterated directional filterbanks. Next, directional energycomponents for each block of the decomposed subband outputs arecomputed. The proposed algorithm captures both local and globaldetails in a palmprint as a compact fixed length palm code of thecomputed directional energies. Palmprint matching is then performedusing Normalized Euclidean Distance classifier. The proposedalgorithm is tested on a total of 500 palm images of GPDS Handdatabase, acquired from University of Las Palmas de Gran Canaria.The experimental results demonstrated the feasibility of theproposed system by exhibiting Genuine Acceptance Rate of 98.2%,Decidability Index of 2.6212 and Equal Error Rate of 0.7082%.