Publication AbstractClose Publication Abstract

Correlation and Classification of Single Kernel Fluorescence Hyperspectral Data with Aflatoxin Concentration in Corn Kernels Inoculated with Aspergillus Flavus Spores.

Yao, H., Hruska, Z., Kincaid, R., Brown, R., Cleveland, T., & Bhatnagar, D. (2009). Correlation and Classification of Single Kernel Fluorescence Hyperspectral Data with Aflatoxin Concentration in Corn Kernels Inoculated with Aspergillus Flavus Spores. ISM Conference 2009 Proceedings. Tulln, Austria: International Society for Mycotoxicology.

Abstract

The objective of the present experiment was to use fluorescence hyperspectral imaging to classify single aflatoxin contaminated and non-contaminated corn kernels. Previous studies showed that visible near infrared hyperspectral imaging technology could offer a novel non-invasive approach toward screening for toxigenic and atoxigenic fungi by identifying a given specimen based on its spectral signature. Additional spectral information may be gained from combining hyperspectral imaging with UV excitation, where the resulting fluorescent image might reveal more information than the reflected image. In this experiment, aflatoxin contaminated corn kernels were produced through inoculation of corn ears in the field with Aspergillus flavus spores. The fluorescence hyperspectral image was acquired with a fluorescence hyperspectral imaging system under a UV light source with 500 ms integration time. Fluorescence image data was first preprocessed through a set up steps. Regions of interest were then created for each kernel to separate corn from background and each kernel was assigned a unique identifier. Spectral signatures and statistical information were extracted from each kernel. Imaged kernels were also chemically analyzed for aflatoxin content. Initially, two-class discriminate analysis models, Resubstitution and Cross-Validation, were used to class corn kernels into two classes: contaminated (ppb level) and control (


Geosystems Research Institute  •  Contact GRI
Modified: December 2, 2020  •  WebMaster  •  Intranet