Trainable algorithm for inspection of soybean seed quality

W. W. Casady, M. R. Paulsen, J. F. Reid, J. B. Sinclair

Research output: Contribution to journalArticle

Abstract

An image pattern classification program was developed to discriminate among asymptomatic soybean seeds, immature seeds, and seeds that had been discolored by fungi or a virus. The program was trainable and could be retrained by the user by recording images of exemplars. The algorithm used chromaticity coordinates to correctly classify asymptomatic seeds, seeds infected by C. kikuchii, seeds that belong to a group used by the Federal Grain Inspection Service called 'seeds of other colors', and 'materially damaged seeds' with 94%, 97%, 85%, and 96% accuracy, respectively. The comprehensive results for all tests yielded a classification accuracy of 94% for classification of seeds into classes that conformed to USDA/FGIS grading procedures. The variables used for classification were color chromaticity coordinates and seed sphericity.

Original languageEnglish (US)
Pages (from-to)2027-2034
Number of pages8
JournalTransactions of the American Society of Agricultural Engineers
Volume35
Issue number6
StatePublished - Nov 1 1992

ASJC Scopus subject areas

  • Agricultural and Biological Sciences (miscellaneous)

Fingerprint Dive into the research topics of 'Trainable algorithm for inspection of soybean seed quality'. Together they form a unique fingerprint.

  • Cite this

    Casady, W. W., Paulsen, M. R., Reid, J. F., & Sinclair, J. B. (1992). Trainable algorithm for inspection of soybean seed quality. Transactions of the American Society of Agricultural Engineers, 35(6), 2027-2034.