Share Email Print
cover

Proceedings Paper

Aflatoxin detection in whole corn kernels using hyperspectral methods
Author(s): David Casasent; Xue-Wen Chen
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Hyperspectral (HS) data for the inspection of whole corn kernels for aflatoxin is considered. The high-dimensionality of HS data requires feature extraction or selection for good classifier generalization. For fast and inexpensive data collection, only several features (λ responses) can be used. These are obtained by feature selection from the full HS response. A new high dimensionality branch and bound (HDBB) feature selection algorithm is used; it is found to be optimum, fast and very efficient. Initial results indicate that HS data is very promising for aflatoxin detection in whole kernel corn.

Paper Details

Date Published: 30 March 2004
PDF: 10 pages
Proc. SPIE 5271, Monitoring Food Safety, Agriculture, and Plant Health, (30 March 2004); doi: 10.1117/12.516135
Show Author Affiliations
David Casasent, Carnegie Mellon Univ. (United States)
Xue-Wen Chen, California State Univ. (United States)


Published in SPIE Proceedings Vol. 5271:
Monitoring Food Safety, Agriculture, and Plant Health
George E. Meyer; Bent S. Bennedsen; Yud-Ren Chen; Shu-I Tu; Andre G. Senecal, Editor(s)

© SPIE. Terms of Use
Back to Top