CLASSIFICATION OF VARIOUS GRAINS USING OPTICAL-PROPERTIES

Rapid identification of the principal constituents of bulk grain sampl es can be used to automate the movement of grain through a terminal or processing plant and can also be important for quality control. This paper examines the potential of using tristimulus values, color solid scale values, and reflectance characteristics of grains as discriminat ing features among various seed types from bulk samples of cereals, pu lses, oilseeds, and speciality crops. Visible light between 400 and 70 0 nm was used to collect data for separation of various grains. Most o f the grains could be distinguished from each other using X, Y, and Z tristimulus values; L, a, and b color solid scale values; or percent r eflectance ratios and ratios of slopes of the percent reflectance curv es at different wavelengths. The bulk reflectance measurement provides a rapid first stage identification and narrows the possibilities for a second, more detailed, identification stage.