Abstract—Palmprint biometric trait is suitable to identify a person as it has more number of features compared to other biometric traits. In this paper, we propose matching level fusion based pamprint identification using WHT and SD. The pre-processing is done on palmprint images to crop ROI and resize ROI into 170×170 size. The Walsh Hadamard Transform (WHT) is applied on ROI of palmprint database and Test palmprint to derive WHT co-efficients which form features. The WHT features of test palmprint are compared with Database palmprint images using Euclidean Distance (ED) to compute EER. The Spatial Domain (SD) features are obtained from ROI palmprint database and Test palmprint. The SD features of test image are compared with features of database images using ED to compute EER. The values of EER computed using WHT and SD features are fused using Log-Transformation to obtain better EER. It is observed that the values of EER are zero in the proposed algorithm compared to non-zero EER values of existing algorithms.

Index Terms—Palmprint, WHT, SD, ED, Fusion

Cite: Chandrakala V., Sandeep B. K., K. Suresh Babu, K. B. Raja, Venugopal K. R., and L. M. Patnaik, "Matching Level Fusion Based PalmPrint Identification using WHT and SD," International Journal of Electronics and Electrical Engineering, Vol. 1, No. 3, pp. 165-170, September 2013. doi: 10.12720/ijeee.1.3.165-170