While long-wave infrared (LWIR) cameras are excellent for detecting thermal patterns around room temperature, they require special, expensive optics made with germanium, sapphire, or silicon
Sensors Unlimited, Goodrich, developer of shortwave infrared (SWIR) imaging systems based on indium gallium arsenide (InGaAs) technology, has recently developed new machine vision cameras for remote monitoring of hot-end processes, such as glass and metal smelting.
How to remotely measure product size, detect process end-points, detect defects, control process temperatures and inspect process furnaces are current challenges that are being met by SWIR imaging technology.
Sensors Unlimited's linear array cameras feature two ranges of sensitivity, from 0.9 to 1.7microns and from 1.1 to 2.2 microns, and the SWIR area arrays with sensitivity ranging from 0.9 to 1.7microns.
There are several distinct advantages to using SWIR-InGaAs cameras.
While long-wave infrared (LWIR) cameras (sensitive from 7 to 14microns) are excellent for detecting thermal patterns around room temperature, they require special, expensive optics made with germanium, sapphire, or silicon.
This elite requirement for LWIR imagers limits the working distance and/or field of view options available to machine vision system integrators.
Sensors Unlimited's SWIR cameras use readily available and inexpensive photographic or video lenses to view hot processes, making it possible for machine vision integrators to use standard glass safety windows in NEMA-rated enclosures, which is especially important when windows are needed to shield equipment from hot splatter.
These windows often need to be replaced, which is cost prohibitive when exotic window materials are required.
In addition, high-end LWIR cameras are very expensive due to cryogenic cooling requirements, while lower cost microbolometer cameras are limited to frame rates of 30 or 60Hz.
Their calibrations also tend to drift which requires a mechanical shutter to create a new non-uniformity correction every few minutes, reducing long-term reliability and disrupting data acquisition of continuous processes.
InGaAs shortwave infrared imagers offer several other key advantages over LWIR technology for machine vision applications.
They hold their factory uniformity corrections for the lifetime of the camera, have no moving parts, can allow for high frame rate imaging to provide 640x512 pixels with 25-micron pixels with frame rates over 100 frames per second (fps); and offer windowing to achieve high frame rates of over 3000 fps for a 64x64 region of interest (SWIR area cameras), High performance InGaAs cameras and arrays see hot spots over 120C in low level ambient lighting, over 150C in room lighting and are regularly used to inspect hollow glass bottle manufacturing lines.
These imaging systems can verify product dimensions and find defects on both the inside and outside of the bottle while monitoring the temperature and its uniformity across the bottle .
In metal processing, separating the good metal from the impurities is a critical process.
In many smelting applications the light intensity changes as the process proceeds, due to differences in emissivity between various materials.
Though this emissivity change is detectable by a variety of camera technologies, some have found the best signal to noise ratio between emissivity differences is in the SWIR waveband.
Industrial process monitoring in the SWIR offers process engineers a great tool to enhance current machine vision systems by imaging beyond what can be seen in the visible wavelength band.
Working closely with the system integrator and camera manufacturer, product quality, consistency and yield can be greatly improved, a sure recipe for improving the organization's bottom line.
Authors: Martin Ettenberg, director of imaging and Douglas Malchow, applications engineer, Sensors Unlimited, Goodrich Corporation.
