Multispectral palmprint identification by Hybrid Haar Wavelet employing multiple score level fusion modus operandi

The Hybrid Haar Wavelet proposed in this paper is simple to generate and exhibit an excellent value of EER as system implementation. The Hybrid Haar Wavelet is generated by reiterated kronecker product of basis haar wavelet with itself. This paper describes a new method to authenticate individuals based on their palmprint identification. ROC plot of FMR v/s FNMR is generated based on the values obtained by score level fusion of red, blue, and green palm scans. The Hybrid Haar Wavelet proposed in this paper provides for a fool-proof system with GAR of 91.2% at a threshold of 7. One-to-many identification of 6000 multi-spectral palmprint images from 500 different palms is used to validate the performance of the system, with 14 fusion schemes proposed to increase GAR while reducing FMR of the system. Minimum EER obtained is 2%, with a maximum GAR of 100%. The system being fool-proof can be efficiently implemented for high security applications like military access.