Giving machines the ability to “see” through cameras and other sensors was a significant milestone for the growth of automation and artificial intelligence. At first, the scientists and engineers developing vision technologies were content with simply trying to mimic the abilities of the human eye. Hyperspectral imaging goes beyond those abilities to capture even more information. Here’s what you need to know about this growing trend in machine vision.
Hyperspectral Imaging Grows Out of Traditional Machine Vision
Most business owners and managers whose facilities use automation are aware of traditional machine vision capabilities. These systems often use optical cameras to capture images, much like a digital camera. The images make use of on-camera, embedded vision systems, or they send the images off to a computer (or the cloud) for processing.
Hyperspectral imaging kicks this process into overdrive. High-resolution spectral data is collected from every pixel in the image. Instead of simply capturing color and light, hyperspectral images use additional wavelengths of light to gather more information for decision-making. Hyperspectral images can even detect differences inside the object, not just on the surface.
How Hyperspectral Imaging Is Revolutionizing Machine Vision
A rapidly expanding group of industries is using hyperspectral imaging already. Similar to the function of a laboratory spectrometer, hyperspectral imaging can be used to detect the chemical content of a product. Pharmaceutical companies can use this technology to detect the wrong pills in a drug packaging operation, potentially saving hundreds or thousands of lives each year.
Food producers can use hyperspectral imaging to help with the sorting of food products. It’s not always easy to tell if a product is spoiled based on its visual appearance. But machine vision systems enhanced with hyperspectral imaging can find evidence of spoilage down to the pixel. And with different wavelengths, it’s even possible to see into fruits, vegetables, and meats to sort out the spoiled foodstuffs.
The medical industry is also finding success with the use of hyperspectral imaging. A simpler patient-friendly tumor screen system uses the technology to detect different biomarkers within a tumor. It’s even able to pull this off at a lower cost than traditional scanning. It’s also more portable. The use of hyperspectral imaging will help improve the care given to patients in rural and remote regions.
Challenges are continuously being addressed and overcome. In the past, the viability of hyperspectral imaging was limited because of the processing speed needed to capture and analyze these data-heavy images. Faster chips, faster sensors, embedded vision, and artificial intelligence have all been able to boost the performance of hyperspectral imaging systems.
Want to add machine vision to your facility? If you’re looking for a quality camera, purchase a camera from Phase 1 camera for your machine vision project!