Title

Maximum discrimination approach for classification of nearly identical signatures

Document Type

Conference Proceeding

Date of Original Version

1-1-2015

Abstract

A method of classifying signatures that are very nearly identical is described. In this instance, slight template errors can result in poor classification performance even in a high signal-to-noise ratio environment. To ameliorate the uncertainty in the signature templates a new approach is proposed. It involves processing only the maximum discrimination frequency bands. A computer simulation using actual chemical signatures illustrates the approach and demonstrates its improved performance over traditional techniques.

Publication Title

Proceedings of SPIE - The International Society for Optical Engineering

Volume

9455

Share

COinS