Accurate seed sorting with machine vision
Agricultural Research Service (ARS) engineer Thomas C. Pearson has developed a highly-accurate portable machine that uses machine vision to sort seeds.
Agricultural Research Service (ARS) engineer Thomas C. Pearson has developed a highly-accurate portable machine that uses machine vision to sort seeds, according to anARS press release.
Pearson had previously developed similar machine vision equipment, but with this seed-sorting model, the work is simpler and faster, according to Pearson. Seeds are placed into a vibrating hopper where they slide down any of the three chutes. Once the seeds fall from the chute, an ARS-developedcolor camera with a CMOS image sensor captures images of the seeds and sends it via a circuit board to a chip for image processing.
The chip, which is equipped with pre-programmed data that determines whether a seed’s surface texture and RGB color values more closely match those of an “accept” seed rather than those of a “reject." Seeds that closely match the parameters defined for “rejects” are quickly dislodged from the machine via an air blast into the reject container, while those seeds that pass fall into the “accept” container.
In Pearson’s tests at his ARS laboratory at theCenter for Grain and Animal Health Research, he showed that the sorter can help wheat breeders to separate kernels of hard red wheat from kernels of hard white winter wheat with 98.6% accuracy. In addition, the sorter was accurate 94% of the time when separating yellow flax from brown flax seed.
View theARS press release.
Also check out:
Package inspection system tracks and traces fresh produce
Robot lettuce pickers and the future of farming
Factory automation market could reach $185 billion by 2016
(Slideshow) Five applications we never thought of for 3D scanning technology