Cereal Chem 66:233-237   |  VIEW ARTICLE

Discrimination of Wheat and Nonwheat Components in Grain Samples by Image Analysis.

I. Zayas, Y. Pomeranz, and F. S. Lai. Copyright 1989 by the American Association of Cereal Chemists, Inc. 

This study describes use of image analysis to discriminate between wheat and nonwheat and between weed seeds and stones in the nonwheat part of a grain sample. Two methods of pattern recognition were used in the discrimination: multivariate discriminant analysis to distinguish between wheat and nonwheat and among weed seeds, and development of a structural prototype to distinguish between wheat and nonwheat. The structural prototype method discriminated well between wheat and nonwheat. Weed seeds were differentiated by multivariate discriminant analysis, but identification of stones posed a problem with this method. Consequently, physical separation of stones prior to the image analysis program described in this paper may be necessary for satisfactory discrimination.