CLASSIFICATION OF CEREAL GRAINS USING MACHINE VISION: I. MORPHOLOGY MODELS

Abstract

An algorithm was developed based on morphological features to classify individual kernels of CanadaWestern Red Spring (CWRS) wheat, Canada Western Amber Durum (CWAD) wheat, barley, oats, and rye. Twenty-threemorphological features were used for the discriminant analysis. Grains from 15 growing regions (300 kernels pergrowing region) across Western Canada were used as the training data set and from another five growing regions as thetest data set. The classification accuracies of individual kernels using the 10 most significant features in the morphologymodel were 98.9, 93.7, 96.8, 99.9, and 81.6%, respectively for CWRS wheat, CWAD wheat, barley, oats, and rye whentested on an independent data set (i.e., the test data set where the total number of kernels used was 10 500; for CWRSwheat, 300 kernels each were selected for three grades). When the model was tested on the training data set (total numberof kernels used was 31 500), the classification accuracies of CWRS wheat, CWAD wheat, barley, oats, and rye were 98.9,91.6, 97.9, 100.0, and 91.6%, respectively.