Classification of edible oils using 532 nm laser-induced fluorescence combined with support vector machine

Abstract

In this paper, laser-induced fluorescence (LIF) is used to characterize and distinguish between different vegetable oils, including soybean, olive, grapeseed, rapeseed, corn, peanut, sunflower, canola, and walnut oils. A 532 nm laser, rather than an ultraviolet (UV) light source, is proposed and used as an excitation light source for the fluorescence analysis of edible oils. It was found that this laser is superior to UV lasers, the fluorescent characteristics become more distinct under 532 nm laser excitation. Edible oils were differentiated by LIF combined with principal component analysis which was used to reduce the dimensionality of data by finding key attributes, and support vector machine. This paper demonstrates, that for ten popular edible oils, the recognition rate can reach up to 100% when a 532 nm laser serves as an excitation light source.