A new approach to a coding and retrieval system for infrared spectral data: The `Effective Peaks Matching' method

Abstract

A new approach to a coding and retrieval system for infrared (IR) spectral data for matching sample spectrum with reference spectrum is proposed. The concepts of “effective peak” and “normalization factor (NF)” are introduced for the coding system. “NF” is related to intensity and is defined as the ratio of the absorbance of the effective peak to the sum of absorbances of all the effective peaks. The algorithm for picking effective peak has been formulated. The new coding system is based on the feature extraction of spectra and both the band position and NF of the effective peak are employed. In the proposed retrieval system, the concept of matching effective peaks of a sample with those of a potential candidate is defined by means of t-test. In a more sophisticated approach, similarity is objectively defined as a membership function, which is equally affected by two membership functions, namely, M(wave) and M(NF), and has a value of 100 for perfectly matched spectra. A list of top scoring compounds is presented in the output. The software developed for the proposed system is particularly suitable for libraries in digitised form and can be used to build proprietary libraries of spectra.