Development of a Software Tool to Perform Qualitative Analysis on Spectral Data Obtained from a NIR-MOEMS Spectrometer

Loading...
Thumbnail Image

Authors

Bajwa, Tejbir S.

Issue Date

2009

Type

Thesis

Language

Keywords

Cluster Analysis , MOEMS spectrometer , NIR spectroscopy , Principle Component Analysis , Savitzky-Golay Differentiation

Research Projects

Organizational Units

Journal Issue

Alternative Title

Abstract

Quality analyses are a vital feature in process monitoring at all production stages. The objective of this thesis is to develop a software tool which can qualitatively analyze spectral data obtained from a Near Infra Red Micro Opto Electro Mechanical Systems (NIR-MOEMS) spectrometer. NIR spectroscopy is a cost effective, non evasive and faster method for analyzing the samples. NIR-MOEMS spectrometer can be used to measure the spectra of samples at any stage of sample manufacturing. Quality analyzing of the samples requires comparing of an unknown sample with samples which are already known. So the software tool developed qualitatively analyzes the data available from the spectrometer at less cost, high speed and using various mathematical operations as pretreatment methods. The software developed requires libraries to be made of known samples and unknown samples are compared with known samples stored in these libraries. Some of the advantages of this software tool over other tools are that this tool can convert the reflection data obtained from NIR-MOEMS spectrometer to absorption data. The software tool is also able to select the wavelength range for comparison depending on user choice provided that this range is a subset of original range obtained from the spectrometer. Libraries of known samples can be made at different wavelength ranges and can be used at any stage of comparison. Future work will focus on development of the tool so that it can be used more dynamically and better methods of data pretreatment are incorporated.

Description

Citation

Publisher

License

In Copyright(All Rights Reserved)

Journal

Volume

Issue

PubMed ID

DOI

ISSN

EISSN