Bit-planes decomposition with eigenpalm on different distance measures

In this paper, a palmprint recognition using bitplane extraction with Principal Component Analysis is presented. Different distance measures are applied for classification to evaluate the recognition performance. Also, bit-plane is selected by analyzing the principal components. Hong Kong PolyU Palmprint Database is applied in this paper. The result showed that the palmprint recognition can achieved 90.42% performance rate by using approximately 4.17% of total principal components by using Manhattan Distance.