 |
23 Nov 2004
Thinking of buying an infrared camera for generating thermal images? Don't forget that its optics and software are just as important as its sensor chip, says Christiaan Maras.
From Opto & Laser Europe December 2004
Infrared thermography - the generation of calibrated thermal images by an infrared (IR) camera - is an invaluable imaging tool for a vast range of industrial and scientific uses.
Applications range from the analysis of hot spots in small electronic circuits to optimizing the performance of rocket engines, for example. The technique's key attraction is that it is non-invasive and can generate an accurate temperature map (thermogram) of an object without the need for any contact.
Thermal images created using this technique allow even the smallest of temperature differences to be detected, which can be vital for spotting flaws in a design or the onset of failure in a product. As temperature patterns can be very difficult to predict, the use of a camera is far more convenient than the alternative approach of attaching thermocouples to the target object. Not only is it sometimes hard to know where to place the thermocouples, but their presence can actually influence the temperature distribution of an object. By contrast, an infrared image will instantly show the presence of hot or cold spots and the measurements can continually be updated in real time.
When an organization commits to purchasing an IR camera it is absolutely vital that it selects a system that is suitable for use with all the conditions that are likely to be encountered. The right camera can provide a fast return on investment. However, it can be quite daunting trying to match a camera's performance parameters to application requirements. Is a camera's dynamic range more important than spatial resolution? Is noise-equivalent temperature difference (NETD) more important than quantum efficiency? Is the camera's frame rate more important than its measurement accuracy? Which type of detector should be used?
Of course, the answers to these questions depend on the intended use of the camera. To help simplify the process, here is a description of the most important camera features (detector type, accuracy, dynamic range, noise, optics and analysis software) that need to be taken into account.
Detector type • Indium antimonide (InSb) detectors High-performance thermal imaging cameras based on cooled indium antimonide focal plane arrays (FPAs) offer excellent sensitivity in the 3-5 μm waveband. Noise levels below 20 mK and the high-speed response of the InSb photodetectors enable high-quality images to be captured over timescales as short as 5 μs. An InSb camera is a good choice when high sensitivity over a broad range of operating conditions is important, and when short integration times or fast frame rates are required. Typical applications for InSb detectors include medical imaging, mid-infrared spectroscopy (analysis of chemicals, rocket and missile exhaust), target signature analysis, astronomy (imaging of molecular clouds and "cold" celestial objects), and thermal imaging of dynamic scenes (propagation of bullets or airbags).
• Uncooled microbolometer These detectors operate in the long-wave infrared band in the 7.5-13 μm range. The advantage of these uncooled detectors is that unlike cooled versions, they do not have any moving parts, reducing the need for maintenance. In addition, uncooled detectors are cheaper to produce than cooled devices. Performance has also significantly improved recently in terms of spatial resolution and temperature sensitivity. Today's uncooled systems can be employed in applications in which only cooled systems could be used previously. The dynamic range of the latest uncooled sensors has also improved and they can now measure temperatures ranging from the subzero to 2000 °C. Most systems allow real-time imaging, but for really high-speed imaging cooled systems still offer an advantage as their time constant is shorter than for uncooled based systems.
• Quantum-well infrared photodetectors (QWIPs) QWIPs operate in a very narrow wavelength band (between 8 and 9 μm), but provide outstanding thermal resolution and unparalleled temperature-measurement capabilities. They are a good solution when ultra-high thermal sensitivity and high-speed data-acquisition is required in the long wavelengths. QWIP detectors operate in a temperature range from -20 to 2000 °C and are able to detect temperature differences as subtle as 0.02 °C.
Accuracy and repeatability The accuracy and repeatability of a camera's temperature measurements are critical factors that should form the foundation of any purchase. However, differences in the performance of various cameras may not be obvious from reading the product literature.
Specifications listed in product data-sheets should be treated with caution, as they often relate to measurements made only at the camera's calibration temperature at the time of manufacture. Checking the temperature-stability of these figures is very important, as a camera's accuracy can change under different conditions. To avoid any doubt, ask the manufacturer about performance under your expected operating conditions.
Dynamic range Dynamic range is the ratio of the highest to the lowest temperature that the instrument will faithfully measure. Again, buyers should be careful, as sometimes this term refers to the inherent properties of the sensor in the camera, while at other times the digitizing capability of the camera electronics is quoted instead. For most users, it is the combination of these two parameters with the final system output that is important. Highly sensitive detector technologies often do not have adequate dynamic range for many applications, which means that they cannot image both ambient temperatures and targets over 70 °C at the same time without saturation. This capability is critical in many electronics applications in which temperature data is taken at regular intervals from the moment when power is applied to a circuit or device.
Infrared cameras equipped with a limited temperature span (dynamic range) will need to be adjusted frequently throughout the data-collection cycle to avoid image saturation. This approach is not practical in many scenarios and often results in measurement errors related to calibration inconsistencies at different camera range settings.
Detector technologies such as quantum-well techniques are much more appropriate for these applications because they maintain high sensitivity over large temperature ranges without imposing a requirement to adjust for saturation.
Noise and sensitivity Noise-equivalent temperature difference (NETD), also commonly referred to as sensitivity, is a measure of an infrared camera's ability to discern small differences in temperature. This parameter is critical when the purpose of the study is to discern very small temperature differences. Observing temperature changes of as little as 0.02 °C can be difficult or impossible with a camera that has a noise threshold of 0.1 °C.
Quantum-well detectors and indium antimonide sensors both produce images with excellent sensitivity. Often there is a trade-off between a camera's sensitivity and the quality of its other parameters, such as dynamic range. As a result, check that the performance of both is appropriate for the application requirement.
Spatial resolution High-resolution (640 x 480 or 320 x 240 pixels) focal plane array detectors are now common in the infrared industry. These are very useful for imaging small features on targets such as printed circuit boards. However, in other, less demanding applications, low-resolution (120 x 160) camera technologies will suffice. The distance from the target and measurement spot-size requirements should be determined to ascertain the importance of spatial resolution.
Optics A sensor may boast good spatial resolution, sensitivity and dynamic range, but these are of little importance if the camera system cannot be equipped with the optics necessary to properly resolve the target. Microscope optics will be required to resolve small targets such as integrated circuits. Without them, the camera is of little value.
Likewise, distant targets will not be resolved without the use of suitable telephoto lens. No degree of pixel magnification will make up for a poor imaging performance. Optical design that eliminates the effects of stray radiation is critical in applications in which targets under test are adjacent to other high-temperature sources.
It is also worth remembering that after several years of ownership, the camera may be required to tackle a new application that has quite different imaging requirements. With this in mind, a unit that can be fitted with a wide array of component options offers greater flexibility that may prove to be beneficial in the future.
Image processing software Most infrared camera users are required to produce data and document their measurements. The ease with which image and temperature data can be extracted from the system can have a dramatic impact on the camera's usability. It is a great advantage if the whole system, from the camera to the post-processing software, uses standard file formats allowing for easy sharing of data. Software that is mated with the camera should ideally support automated transfer of temperature data to a variety of other programming environments such as Microsoft Excel or Matlab. The supplied software should support exportation of test sequences to common formats such as .AVI that can be embedded into presentations.
Cameras and camera software that take into account the calibration effects of additional optics or window materials are critical for accurate temperature measurements.
The researcher should also have flexibility in choosing the host computer and operating environment. This means that software specifically designed for use with the infrared camera should be supported on a variety of operating systems such as Windows 98, Windows 2000, Windows XP and Windows NT. This allows the researcher to use the computer that is most appropriate to host not just the camera's image-processing software, but other software needed to support data-collection and processing.
Conclusion An infrared camera is a long-term investment, so you should deal with a manufacturer with a reliable background that will still be trading in the future. Go for a supplier that can offer you a choice of different systems, detectors and integration times, and avoid those that only offer one particular system. They may want to force you into buying things that you do not really need.
See to it that the manufacturer can support your future requirements. A wide range of accessories such as lenses and filters should be part of a good manufacturer's offering. And finally, see to it that your supplier is able to give you the necessary infrared training, in-house support and service on a local level. Communicating with someone who speaks a different language and is thousands of kilometres away when you have a problem is not the easiest option. Neither is sending your camera to the other end of the world when it needs to be serviced or calibrated.
 |  |  |  |  |  |  |
| © 2024 SPIE Europe |
|
