SWIR Imaging in food safety processes
Introduction
Imaging the reflectivity or transmissivity light, from or through materials and the signature a material leaves on the various wavelengths imaged, has become an important factor in the advancement of the machine vision industry. By utilising different wavelengths of light, machine vision systems can perform a variety of inspection tasks with exceptional precision and efficiency.
For example, ultraviolet (UV) light, which has shorter wavelengths than visible light, can be used to inspect materials for cracks and other types of damage. UV light is absorbed by most materials and only penetrates short distances, making this an ideal tool for surface inspection. On the other side of the visible light spectrum is the infrared (IR) spectrum, which starts just beyond what we see as red visible light and extends all the way through the electromagnetic spectrum to what we feel heat (thermal energy), also known as long-wave infrared light (LWIR).
Please note that for the purposes of this article we are defining Near Infrared light (NIR) to have wavelengths from 700nm to 1000nm and Shortwave Infrared light (SWIR) to range from 1000nm to 3000nm. These two wavelength range definitions coincide with the sensor material used to image them. Silicon is sensitive through the visible and up to 1000nm and InGaAs and/or MCT from 900nm to 3000nm. In other circumstances we may define NIR as the range from 700nm to 1700nm and SWIR from 1700nm to 3000nm. The name is not important, it’s the wavelengths at which a camera is sensitive that define its capabilities.
Wavelengths from slightly longer than visible light (700nm) to the longest short wave infrared (3000nm) can be commonly used in machine vision applications. For example, NIR and SWIR Infrared imaging are commonly used in agricultural or food applications: detection of crop stress, soil nutrients, fruit grading, food and beverage manufacturing, contamination detection, other quality issues and more.
For materials such as plastics, ceramics and semiconductors that are difficult to observe in the visible spectrum, the longer wavelengths from 900nm to 1700nm (InGaAs sensors) provide significantly better inspection capabilities than visible or short NIR wavelengths. Different plastics for example can be sorted from each other with these wavelengths because their different molecular structures interact with wavelengths in this range and so produce individual signatures in their reflected spectrums.
Image 1: The light spectrum consists of a range of wavelengths, each with its own set of properties for a wide range of machine vision inspection tasks.
Adoption of SWIR imaging technology in the food and beverage industry
The food and beverage industry has strict quality control standards in place to ensure that consumers receive safe, high-quality products. With its expanded inspection wavelength range, InGaAs sensor cameras offer faster, more accurate and reliable results, enabling more advanced inspections that result in less contamination, improved sorting capabilities and higher quality products. In recent years, SWIR technology has gained popularity in the food and beverage industry thanks to advancements that have enhanced image quality and affordability.
Grain inspection
Bulk grains such as rice and wheat often contain varying degrees of impurities, including foreign materials and discoloured grains. SWIR imaging can be used to classify and separate objects such as grains, nuts and corn, based on size, shape and other factors to accurately identify contamination and defects for efficient sorting.
High precision and Real-time inspection
Cameras such as JAI’s SW-4010Q linescan camera, capable of providing simultaneous RGB and SWIR imaging, support high-resolution image capture of granular materials in a single pass with pixel-level correlation and no parallax issues. This enables precise inspections, making it possible to identify even the smallest defects with precision. The incorporation of SWIR technology allows for real-time inspection at high speeds, enabling immediate corrective action and reducing the resources required for rework.
Image 2: Detection of impurities in red bean inspection using SWIR plus RGB imaging
Deeper penetration and Moisture Analysis
SWIR cameras are better suited for analysing dense and opaque materials that require deeper penetration. SWIR wavelengths penetrate the outer layers of grain kernels to detect moisture or other properties inside, can be used to assess grain quality and nutritional value. Additionally, because water molecules absorb light at specific wavelengths in the SWIR spectrum, cameras can generate moisture maps to identify excessive hydration, which can lead to spoilage, or insufficient hydration, which causes brittleness.
Image 3: Using SWIR to reveal foreign objects when inspecting nuts
Packaging inspection
Packaging inspection is a comprehensive process used to verify quality and safety requirements. Using cameras integrated with SWIR technology offers several functional benefits. SWIR light can pass through many types of plastic and the surface printing layer, providing clear images of contents within opaque containers to verify fill levels. Furthermore, different materials like plastic, paper and metal exhibit different reflectance patterns in the SWIR region. This allows manufacturers to confirm that paper-based packaging is of consistent quality and not contaminated with harmful substances that can leach into the product.
For packaging defect detection, spectral transmittance in the SWIR waveband allows for the identification of oils, grease and air pockets. For example, systems can detect oil contamination on cans or identify weak seals in laminated films and pouches that can lead to spoilage or contamination.
Integration of RGB and SWIR Sensors
While visible light cameras are necessary to analyse colour nuances, size and shape, combining them with SWIR imaging in a single camera device provides significant mechanical advantage and removes the issues with calibration, occlusion and parallax when using two cameras.
Better quality control
Simultaneous use of RGB and SWIR imaging provides a more comprehensive analysis of food products. RGB imaging captures visible light to detect surface defects and colour variations, while SWIR imaging captures light beyond the visible spectrum to penetrate certain materials and detect subsurface or spectral absorption features. This combined data enables more accurate and reliable inspection to meet product quality and safety standards.
Image 4: Simultaneous inspection of both the colour printing and contents of a package with RGB+SWIR technology.
Reduced complexity and efficiency
By combining two imaging techniques into one device, the need for separate cameras, lenses and cabling is eliminated. This lowers the overall cost and simplifies the integration of the system into an existing production line. Multi-sensor models like JAI's Sweep+ Series provide simultaneous imaging that detects both external discolouration and internal bruising, facilitating the rapid removal of defective products from the processing line.
Summary
SWIR imaging is a valuable technology for the food and beverage industry. With its ability to detect invisible defects and contaminants, SWIR technology helps companies maintain high standards of product quality and ensure products are safe for consumption
Learn more about SWIR imaging for food safety
SWIR imaging transforms food and packaging inspection by imaging materials to detect moisture, internal defects, and contaminants. Integrating SWIR imagers alongside RGB cameras in the same system ensures comprehensive quality control and production efficiency.