John Deere Harvester Works Product Engineering developed an imaging system to measure relative amounts of mechanical damage in corn samples. The imaging system detected the exposed white starch of broken corn kernels. A “picture” of a corn sample was analyzed to determine the percentage of area that was exposed white starch. This percentage correlated with damage values obtained using hand count methods.
The system was non-contacting and portable. It consisted of a personal computer, a video camera, a video signal digitizing card, controlling software, a blacklight, and a computer controlled mixing device.
This paper describes development of the system, the type of hardware selected, how the system worked, correlation to hand counting and some limitations of the system.