Tech Focus: SWIR imaging enriches machine vision
Short-Wave Infrared (SWIR) imaging is a relatively new imaging technology. It captures wavelengths of light within the infrared band of the electromagnetic spectrum that are beyond the visible spectrum. Traditional visible light cameras can only “see” what the human eye perceives, i.e. light between the wavelengths of 400 and 700 nm. However, SWIR sensors are sensitive to infrared light in the 900 nm to 1700 nm range (and even as wide as 700 – 2000 nm), revealing details that are otherwise invisible. This ability to “see” in the SWIR range enables new imaging capabilities, especially in machine vision applications, where precision and the ability to detect subtle features are crucial.
Development of SWIR imaging technology
The study of infrared light began in the early 19th century, long before the concept of machine vision or imaging systems as we know them today. In 1800, British scientist William Herschel discovered infrared radiation while experimenting with the spectrum of sunlight. Using a thermometer, he noticed that temperatures increased just beyond the red end of the visible spectrum, which led to the identification of infrared radiation.
During World War II, infrared technology received a significant boost as the military began to explore ways to detect heat signatures and improve night-time visibility. These efforts led to the creation of mid-wave infrared (MWIR) and long-wave infrared (LWIR) sensors, which are used in military applications like night-vision goggles and heat-seeking missiles.
The SWIR imaging range gained more interest in the 1960s and 1970s, with the development of the InGaAs (Indium Gallium Arsenide) sensor. InGaAs had a significant advantage over earlier infrared materials because it was sensitive to the SWIR range, making it ideal for a variety of applications, including machine vision.
In the 1980s, InGaAs detectors began to be used more widely in scientific research, particularly in remote sensing, astronomy, and spectroscopy. These early SWIR sensors were costly and bulky, but they demonstrated the potential of imaging technology that could “see” beyond visible light.
Later advancements in sensor technologies enabled the miniaturization and commercialization of SWIR imaging systems. The development of solid-state detectors made it possible to build smaller, more affordable SWIR cameras. At the same time, improvements in optical coatings and infrared optics enhanced the performance of these systems, making them more reliable and efficient.
More recently, advances in CMOS (Complementary Metal-Oxide-Semiconductor) technology have facilitated the creation of SWIR cameras with lower power consumption, faster processing speeds, and enhanced image quality. These cameras are increasingly used in environments with challenging lighting conditions or where traditional visible-light imaging is insufficient. The integration of machine learning and AI with SWIR imaging has further enhanced its capabilities, allowing for faster and more accurate analysis of complex data. For example, machine learning algorithms can now analyze SWIR images in real-time, detecting defects and anomalies in industrial products with greater precision than ever before.
Applications for SWIR Imaging in Machine Vision
SWIR imaging has a broad range of applications across various industries due to its ability to capture detailed information that is invisible to the human eye.
SWIR imaging is used extensively in quality control for manufacturing processes where it can detect surface defects, moisture content, contaminants, and other imperfections that are not visible with standard visible light cameras.
For example, in food processing, SWIR imaging can help identify defects such as under ripeness in fruit or contamination in packaged goods. This is because ripe fruits, which typically have higher moisture content, will appear brighter compared to unripe fruits, as more water reflects more light. Contaminants with a different composition to the food around it will also show up clearly in SWIR images. Lynred explain these possibilities in these more detail, underlining how SWIR imaging can be used to improve quality control in food and beverage production.
This feature of differentiating materials based on their unique absorption and reflectance properties means that many materials that appear identical in visible light – such as plastics, fabrics, or metals – have distinctive signatures in the SWIR range. Therefore, in recycling and material sorting, SWIR cameras can be used to accurately identify and separate different materials like paper, plastic, and metal, even when they are mixed together. This makes SWIR imaging an ideal tool for automating recycling processes. We’ve previously reported on the use of SWIR imaging in recycling, covering both SWIR and NIR applications, and TOMRA have further widened their portfolio of SWIR tools since the publication of that blog.
SWIR cameras can penetrate fog, smoke, and dust, and still provide clear images under these conditions, unlike visible light cameras. This makes SWIR an ideal tool for detecting suspicious activity in low-visibility conditions. It is used effectively in border security, airport surveillance, and military applications, where visibility can be impaired by weather or other environmental conditions. Sensors Unlimited, part of Collins Aerospace, explain in their blog why SWIR is a particularly valuable technology for Intelligence, Surveillance, and Reconnaissance (ISR).
In medical research, SWIR imaging is being implemented to visualize tissue structures, blood vessels, and other biological features that are not identifiable in the visible spectrum. SWIR light can penetrate deeper into tissues compared to visible light, making it useful in non-invasive imaging techniques. SWIR imaging can be used in medical diagnostics, such as monitoring blood oxygen levels or detecting skin cancer at an early stage. Researchers at the Wellcome/EPSRC Centre for Interventional and Surgical Sciences (WEISS) at UCL and surgeons at Great Ormond Street Hospital (GOSH) have pioneered SWIR fluorescence for combatting neuroblastoma. Chemicals injected into the bloodstream attach themselves to cancerous cells, and surgical teams using a SWIR camera can image these chemicals and penetrate visually deeper into the tissue to provide sharper, more detailed images. This non-invasive method highlights tissue that needs to be removed and can even be used live during procedures.
SWIR imaging is a powerful tool that is transforming the landscape of machine vision and wider imaging applications. Its ability to provide insights beyond the visible spectrum makes it an indispensable technology for a variety of industrial, medical, and security applications. From quality control and material sorting to night-time surveillance and non-destructive testing, SWIR imaging is helping industries make smarter, more informed decisions while improving accuracy and efficiency. As the technology continues to evolve, we can expect even greater advances in its capabilities, opening up new possibilities in the world of machine vision.
