Colour fusion in multi-spectral face authentication systems

In this paper, the problem of multi-spectral face authentication is considered. The verification process is based on the Normalised Correlation measure within the LDA feature space. First, considering visible and thermal Infra Red (IR) images, different colour transformations are applied and the performance of the system within each colour space is evaluated. The scores associated to an adaptively selected subset of the colour based classifiers are then fused in the decision level. The selection process is based on a sequential search technique called the “plus L and take away R” algorithm. The sum rule is used for fusing the related scores. Our extensive experimental studies using the UTK-IRIS face database demonstrate that using the proposed method, the performance of the system considerably improves as compared to the individual Visible-based or IR-based face verification systems. The proposed multi-spectral colour fusion scheme also outperforms the best colour space in different conditions within the IR and Visible colour subspaces.