An automated seed presentation device for use in machine vision identification of grain
Authors: D.S. Jayas, C.E. Murray and N.R. Bulley
Description: A kernel positioning system was designed and fabricated to automatically pick up and separate kernels of various grain types. The system was tested for its ability to separate kernels of four grain types: wheat (cv. 'Katepwa'), barley (cv. 'Manley'), canola (cv. 'Tobin'), and lentils (cv. 'Eston'). Wheat and barley were tested at three different moisture contents and canola was tested at two moisture contents. To determine if the system had any bias to the selecting of certain seeds, mixtures of different grain types were also used in the tests. The system successfully picked up and separated kernels of wheat, barley, canola, and lentils. The system's ability to pick up and separate kernels was not influenced by moisture content. In mixtures of grains (e.g., barley in wheat at I, 3, and 5% levels by mass), there was no significant difference in the number of imageable wheat kernels. The system, however, had a bias to pick more kernels of the major grain component present in the mixtures of wheat and canola.
Citation: D.S. Jayas, C.E. Murray and N.R. Bulley 1999. AN AUTOMATED SEED PRESENTATION DEVICE FOR USE IN MACHINE VISION
IDENTIFICATIO OF GRAIN. Canadian Agricultural Engineering 41(2):113-118.
Publisher: Canadian Society for Bioengineering
Language 1: en
Rights: Canadian Society for Bioengineering