A Novel Efficient Classwise Sparse and Collaborative Representation for Holistic Palmprint Recognition

Palmprint recognition is an important and widely used modality in biometric systems. It has a high reliability, stability and user acceptability. Although the discriminative ability of the existing state-of-the-art holistic techniques, their effectiveness heavily relies upon the quality of training data. Indeed, palmprint images contain different information including identity, illumination and distortions related to the acquisition systems. To overcome this problem, we explore a novel efficient holistic Classwise Sparse and Collaborative Representation (CSR). Extensive experiments have been performed on two existing and widely used palmprint datasets: multispectral and Poly U. The obtained experimental results demonstrated very encouraging performances when compared to state-of-the-art techniques.