Cereal Chem 66:466-470   |  VIEW ARTICLE

Image Analysis and Characterization of Cereal Grains with a Laser Range Finder and Camera Contour Extractor.

C. Chen, Y. P. Chiang, and Y. Pomeranz. Copyright 1989 by the American Association of Cereal Chemists, Inc. 

An image analysis system with the capability to acquire and combine three-dimensional laser range data and two-dimensional camera contour extracted images was used to discriminate between cereal grains and weed seeds and between soft white and club wheats, two- and six-rowed barleys, and rye and triticale kernels. Discriminant functions were established to discriminate between wheat and nonwheat with 99.5% accuracy and between nonwheat and wheat with 98.3% accuracy. The discrimination between two soft white and two club wheats was about 90% accurate. The discrimination between two- and six-rowed barley was about 90% accurate, between barley and other grains 97% accurate, and between rye, triticale, and other grains 90% accurate. Discrimination between wheat, soybeans, oats, wild oats, dent corn, flint corn, wild buckwheat, sorghum, and lamb's-quarter was almost 100% accurate.