Skip to main content
ARS Home » Research » Publications at this Location » Publication #118795


item Kim, Moon
item Chao, Kuanglin - Kevin Chao
item Chen, Yud
item Chan, Diane

Submitted to: Society of Photo-Optical Instrumentation Engineers
Publication Type: Proceedings
Publication Acceptance Date: 12/28/2000
Publication Date: 12/28/2000
Citation: N/A

Interpretive Summary: Large numbers of people each year are affected by foodborne illness. Unpasteurized apple juice or cider, a major beverage for children in the U.S., has been identified as a repeated source of E. coli O157:H7 contamination. Sources of pathogenic E. coli contamination include apples with physical damage such as legions and bruises, and those contaminated with animal fecal matter. Fecal contamination may occur when bird dropping land directly on apple surfaces, or when apples come in contact with ground where bird droppings or cow manure have fallen. The pathogenic bacteria may also be passed to apples from bites on the apples by an animal that has pathogenic bacteria in its mouth. Recently, we developed a laboratory-based hyperspectral imaging system capable of both reflectance and fluorescence techniques. Hyperspectral imaging methods combine the features of imaging and spectroscopy to acquire both spatial and spectral information from a biological object simultaneously. The objectives of this study were to evaluate fecal contaminated apples using hyperspectral imaging to determine spectral features of normal and contaminated parts, and to further evaluate multispectral approaches using the wavelengths determined for detection of fecal contamination on apples. A nondestructive imaging algorithm for the detection of fecal contamination was developed. Ultimately, a common aperture multispectral imaging system can be developed for a rapid on-line implementation. This research benefits food processing industries and consumers by providing a rapid means to detect unwholesome food products.

Technical Abstract: We examined fecal contamination on apples, as part of on-going food safety research, with the use of the recently developed hyperspectral imaging system that has a spectral range in the VIS to NIR region of the spectrum from 400 to 900 nm. Both reflectance and fluorescence techniques for detection of exogenous fecal contamination on four apple varieties, Red Delicious, Gala, Fuji, and Golden Delicious were evaluated. Thick patches and thin, transparent smear spots of fresh dairy cow manure were placed on these apples to simulate fecal contamination. In addition, these spots were created on sun-exposed side and shaded side surfaces to account for natural color variations due to environmental growth conditions and ripeness. Spectral features from both reflectance and fluorescence spectra of samples including fecal contaminated spots were evaluated to determine wavelengths where minima, maxima, and plateaus occur. Images at these wavelengths were used to create combinations of simple two band ratios, second differences, normalized differences, and absorption depth images. Preliminary results of these simple multispectral approaches indicated that the reflectance method can differentiate thick patches of manure from regions of normal apple surfaces using a two NIR band ratio (R850/R800) with a simple threshold. However, for the detection of thin manure spots, the reflectance method may require more complicated image processing approaches. Fluorescence techniques with a simple two band ratio (F680/F450) differentiated normal apple surfaces from contaminated spots regardless of apple skin coloration and thickness of manure treatments. These results will be further refined to develop a rapid on-line multispectral detection system.