الفهرس | Only 14 pages are availabe for public view |
Abstract Spectroscopy is the study of the interaction between matter and radiated energy. Any interaction comprises radiated energy as a function of wavelength or frequency. This thesis deals with the spectrum of the radiated energy. It considers a type of spectroscopy called Raman spectroscopy as it can be used with solids and liquids with no need for sample preparation. It is non-destructive, and it is acquired quickly within seconds. The thesis studies the applications of Raman spectroscopy and how to use it as an analytical tool. This thesis considers the problem of noise reduction and trend removal from Raman spectra. For trend removal, a proposed method comprising local variance estimation and cubic spline interpolation is presented and tested with different types of trends. Several de-noising methods will be considered in this thesis revealing that wavelet denoising will achieve the most superior results before trend removal. |