Detection using correlation bound in a linear mixture model

A detection approach, detection with a correlation bound (DCB), is introduced based on a linear mixture model. We use the upper bound of the correlation between the target and mixing components as the detection index, and derive the expression for this correlation bound using the observed data. The proposed method is an unsupervised approach and provides satisfactory detection performance without imposing strong assumptions. Simulation and experimental results are presented to demonstrate the effectiveness of the proposed method in Raman spectroscopy for the detection of surfacedeposited chemical agents. r 2006 Elsevier B.V. All rights reserved.