Submitted to: Principle Component Analysis
Publication Type: Book / Chapter
Publication Acceptance Date: October 28, 2011
Publication Date: March 7, 2012
Citation: Liu, Y. 2012. Principal component analysis in the development of optical and imaging spectroscopic inspections for agricultural / food safety and quality. In:Sanguansat, P. Principle Component Analysis. Rijeka, Croatia: InTech. p. 125-144. Interpretive Summary: Concurrent improvements in analytical techniques and data processing enable the spectroscopic devices based sensors to be more sensitive and selective, smaller, cheaper, and more robust than their laboratory version. Both optical and imaging spectroscopy are making critical judgment in assessing the safety, security, and quality aspects of agricultural and food products. To design rapid optical and imaging sensing systems in addressing the urgent need for public health, several essential spectral bands (usually two or three) are desirable. The selected wavebands should not only reflect the chemical /physical information in samples, but also maintain successive discrimination and classification efficiency. This Chapter only outlines the usefulness and effectiveness of PCA in extracting valuable information from optical and imaging spectroscopy of complex agricultural /food matrixes and subsequently the development of optical and imaging spectroscopic tools for their safety and quality assessment within the recent ten years.
Technical Abstract: The Chapter reviews the recent developments of PCA in optical and imaging spectroscopy for agricultural and food safety and quality. Food safety is one of most important issues for public health, and authorities have zero tolerance performance standards for various food products. Driven by this increasing interest of protecting food supply, the ability of optical and imaging spectroscopic techniques to rapidly, routinely, potentially to be portable and on-site, as well as non-destructively screen agricultural commodities sets them apart from traditional analytical or inspection methods that are labor intensive and time-consuming. Despite the great progress in designing next-generation inspection tools, difficulties still exist. One of challenges is how to extract useful and effective information from complicated agricultural and food matrixes. Hence, it needs more focused and detailed study of data mining on these matrixes with the aid of such advanced multivariate data analysis as PCA.