Hybrid DWT-DCT based method for palm-print recognition

Human palm print is a wide spreading biometric characteristic that has been used to detect an individual identity. In this paper, hybrid Discrete Wavelet Transform - Discrete Cosine (DWT-DCT) Transform technique is proposed for the feature extraction in palm-print image. A palm-print image is first segmented into thirty two bands in spatial-domain and then a frequency-domain transformation is applied on each band in order to generate the dominant features containing high compaction energy. The transformations such as Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and proposed hybrid DWT-DCT are used for the purpose of palm-print feature extraction. The Euclidean Distance is used for matching the query image with the images stored in the IIT Delhi Touchless Palm-print database. The Genuine Acceptance Rate (GAR) is applied for the measurement of accuracy. The result shows that the proposed Hybrid method provides better palm-print recognition accuracy in comparison to DWT and DCT.