The importance of spatial resolution in IR thermography temperature measurement: three brief case studies

Abstract

This paper discusses the differences in temperature measurement among three infrared (IR) cameras with different spatial resolutions. The targets used were actual field problems. This was not a laboratory experiment, but a real-world test of these cameras. Images were taken under the same conditions at the same distance with all the IR cameras. The results show there can be dramatic differences in temperature readings illustrating the importance of adequate resolution when making temperature measurements with IR cameras. INTRODUCTION Many modern IR cameras have built-in cross-hairs that show the user the size of the spot being measured. One would think this would solve the problem of measurement error due to a target at too great a distance for its size and IR camera resolution. But in our haste as thermographers under pressure to get results, we often don't take the time to ensure sufficient target size, or we use a function such as the maximum temperature in a box that does not show whether the target is adequate for measurement. Or, we don't have adequate optics or IR camera resolution to get a good measurement of a distant target that just isn't quite large enough. Whether the measurement errors resulting from inadequate spatial resolution really matter depend on the magnitude of the error and the criticality of the target. In our study we found a critical problem on a substation transformer at 40% load that could certainly devolve into a major failure as the load increased to 100% with increasing demand. The question arises as to how much difference this makes anyway. The answer isn't that simple. The degree of measurement error depends on target temperature, size, distance, IR camera and what is behind the target. The worst case scenario is found in outdoor electrical applications where there are numerous targets a few inches or smaller in size that can be 25 to 80 feet distant with no opportunity for the thermographer to get closer. Frequently, behind these targets is a clear, dry sky whose temperature can approach -50°F or colder. The IR camera averages what it captures for a given measurement spot size. When you average in -50°F you can get significant measurement errors even for hot targets that are just a bit too small for the IR camera and optics in use. Adequate IR camera resolution is a combination of the IR camera detector and optics. The wise thermographer ensures the right system is available for the job at hand. This paper illustrates with real world measurements on actual thermal anomalies found in the field, the magnitude of error one can expect with 3 different resolution IR cameras using their standard optics. All measurements were made at about the same time from the same distance on the same targets under the same environmental and load conditions using different models (shown in parentheses below) of FLIR Systems cameras. For brevity, we will call them low, medium and high resolution IR cameras. The low resolution camera has 160 horizontal by 120 vertical pixels (E65), the medium resolution camera 320 horizontal by 240 pixels (P65HS) and the high resolution camera 640 by 480 pixels (P640) in their respective detector arrays. Also for brevity, we use resolution for spatial resolution in this paper.