In simple terms, machine vision refers to the use of cameras, computers, and supporting technologies for the automatic extraction and analysis of information from digital video and images. Often, machine vision is associated with industrial use, including manufacturing applications such as automotive, food and beverage, and semiconductors and electronics, where machine vision performs gauging, identification, guidance, and inspection tasks.
While machine vision has never been limited only to industrial deployments, the technology has long been associated with such applications. As machine vision capabilities grow, costs decrease, and user demands grow, however, its uses continue to expand further beyond the factory floor and into everyday life, perhaps obscuring—to some extent—the meaning of machine vision itself.
Examples of machine vision—or computer vision, if you strictly adhere to more traditional definitions—are well documented, as cameras, vision systems, and computer vision techniques are used in things like automatic license plate readers, virtual and augmented reality setups, robotic vacuums, hobbyist drones, and in smartphones and smartphone apps. Snapchat, for instance, uses various computer vision techniques in its numerous filters or lenses – which first recognize a person’s face, and then make various and oftentimes humorous changes to a person’s appearance and features in the app. Here are some examples of such filters, featuring my sons, Jacob and Brandon.
In this issue, we highlight a few more examples of imaging beyond the factory floor. An article on page 12 describes the design and installation of two imaging chambers at Purdue University, in which multispectral and hyperspectral cameras aid in digital phenotyping.
An article on page 8 also describes the design and development of an automated access control system designed primarily for the workplace which leverages infrared cameras, artificial intelligence, and 3D data to enable frictionless doorway entry. Even further beyond the factory floor is a 3D laser scanner system targeting underwater inspection applications, also on page 8. Designed by Kraken Robotics, this system is designed for the inspection of subsea assets including cabling, ship hulls, anchor chains, and pipelines, among others.
Another article, on page 17, details how two different schools installed video surveillance systems, and how one school succeeded while the other failed as a result of lens selection.
This is all to say that, while market data may indicate that most sales come from industrial applications, machine vision technology adoption continues to grow worldwide, and we plan to cover as much of these emerging areas as possible, going forward.