Abstract
Cross-correlation technique is effective as an alternative to coding procedures for systematizing the manipulation of a large file of digital spectral data. As a search method, it is reliable in the recognition of matching spectra and yields a low incidence of false identifications. Precision is governed by the purity of specimens rather than instrumental factors. The utility of a file is augmented by treatment of the spectra as vectors that are normalized to unit length before storage. Such reduction minimizes computer time required for the determination of correlation coefficients; in addition, proportionality factors are eliminated and plots can be superposed on a concentration-independent basis to supplement correlation searches.
Get full access to this article
View all access options for this article.