ARS | Publication request: MULTISPECTRAL LASER INDUCED FLUORESCENCE IMAGING TECHNIQUES FOR NONDESTRUCTIVE ASSESSMENT OF POSTHARVEST FOOD QUALITY AND SAFETY

Submitted to: Acta Horticulture Proceedings
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: May 1, 2005
Publication Date: June 30, 2005
Citation: Kim, M.S., Lefcourt, A.M., Chen, Y.R. 2005. Multispectral laser induced flourescence imaging techniques for nondestructive assessment of postharvest food quality and safety. Acta Horticulture. 682:1379-1386.

Interpretive Summary: Our laboratory has been developing nondestructive sensing methodologies to address the safety and quality issues related to animal fecal contamination of food products. The goal of image-based (machine vision) applications for real-time food safety inspection, within the confines of a specific sensing method (e.g., fluorescence), is to find the simplest approach that allows processing input data at commercial line speeds. In general, this implies that less computationally complex methods are preferred. In this regard, use of a single fluorescence emission band that allows accurate discrimination of animal feces contamination on apples is ideal. We recently developed a multispectral laser induced fluorescence imaging system capable of ns-scale characterizations of fluorescence decay. The objective of the study was to investigate ns-scale fluorescence emission decay characteristics of apples artificially contaminated with cow feces to determine a suitable red fluorescence band and optimal gate-delay time for detection of fecal contamination on apples. Four spectral bands, F670, F680, F685 and F730 centered near the emission peak wavelengths of the major constituents responsible for the red fluorescence emission from apples and animal feces were examined. The results based on the nanosecond decay curves demonstrated that 670 nm with 10 nm FWHM provided the greatest difference in time-dependent fluorescence responses of feces treated and untreated apples surfaces. This information is useful to engineers and scientists developing fluorescence-based machine vision inspection systems for fecal contamination on agricultural products.

Technical Abstract: Nanosecond-scale, time dependent fluorescence emission characteristics of apples contaminated with a range of diluted cow feces were acquired using a recently developed multispectral laser-induced fluorescence imaging system. Four spectral bands, F670, F680, F685 and F730, centered near the emission peak wavelengths of the major constituents responsible for the red fluorescence emissions from apples and animal feces were examined. The objective of the study was to investigate ns-scale fluorescence emission characteristics of samples to determine a suitable single red fluorescence band and optimal gate-delay time for detection of fecal contamination on apples. The results based on the ns decay curves show that 670 nm with 10 nm FWHM provides the greatest difference in time-dependent fluorescence responses of feces treated and untreated apples surfaces at a gate-delay of 4 ns from the laser excitation peak. With these detection parameters, fluorescence emissions from the diluted feces spot account for a minimum of 27% of the total emissions.