The role of computer vision in producing perfect pasta

Spaghetti harvesting? Did our April Fool’s post catch you out? Of course this story was just a prank but our autofocus-zoom cameras really can be used in agritech, hyperspectral imaging and food production applications.

AND the tree is a real one, the image shows the seed pods of the Catalpa ovata tree. This tree originates from China and is now found widely across North America and Europe. It’s a deciduous tree which blossoms with bell-shaped white flowers in late spring before these bean-like seeds grow, but, sadly, no spaghetti.

*******

We’ve written several posts about the use of computer vision in agriculture, from precision agriculture to keeping an eye on livestock. Pasta producers in the Italian-speaking Ticino region of Switzerland are the latest farmers to employ advanced imaging to ensure a perfect spaghetti harvest.

Growing spaghetti in Ticino dates back to the early 20^th century, and producers have traditionally relied on manual inspection to check the ripeness of the crop. This BBC documentary from 1957 gives great insight into the process, and the historic reliance on a labor-intensive human workforce.

Hyperspectral imaging is now adding a modern touch to this traditional industry. Agricultural robots rigged with cameras with hyperspectral sensors and onboard processing are being trialed to move between the spaghetti trees and image the growing pasta.

The imaging system covers two spectral ranges (380–1030 nm and 874–1734 nm). Pattern texture features including contrast, entropy and homogeneity are extracted from the images at optimal wavelengths, and matched against known data for perfectly ripe pasta. In this way, farmers can be sure of harvesting the crop at the best possible moment. Various robotic grippers are in the test stage as developers search for the perfect robots to pick the spaghetti and pack it without damage, further reducing the need for human pickers.

Such imaging systems require high-quality image capture and processing, view our range of imaging products and contact us with your agricultural, industrial, surveillance or scientific imaging requirements to see which of our products will support your application.