There are many types of grains, such as wheat, barley, rye, rice, etc. And to delve deeper into the matter: there are all sorts of wheats, etc. growing in different countries under different conditions.
And also, different wheat cultivars can occasionally become mixed up during harvest, but breeders and seed companies rely upon pure lots. The grains therefor have to be sorted and this is not an easy task. Differences can be very subtle and the grains are very small. Also, wheat tends to be processed in large quantities and human inspection would be much to slow and extremely expensive.
And therefore machine vision systems come into play, such as the EyeVision software on smart cameras of the EyeCheck 7xxx series. Of course the sheer volume of grains that must be inspected to keep up with production rates is a difficult challenge for most computer vision inspection systems.
One possible solution for the grain industries is a camera that has both, a FPGA (field-programmable gate array) and a CMOS sensor, such as the EyeCheck 7800.
The EyeCheck 7800 smart camera not only has a FPGA of 85K but also a Dual Core processor and a 4.2 MegaPixel CMOS sensor. In combination with the EyeVision 3.0 software it is possible to capture extremely small objects in highspeed. Thanks to the FPGA the image capture and image evaluation can be realized in real time.
Because the inspection programs, created with the EyeVision software, are hard-coded into the FPGA, it is extremely fast and robust to the strains of processing large amounts of data over a lengthy period of time. Additionally the FPGA can be programmed with graphical tools.
This is all in line with the EyeVision image processing software, which is also graphically programmable. The easy-to-handle drag-and-drop programming makes it easy to create inspection programs, even without programming skills.
Source: http://www.evt-web.com/