Back to Basics

[ click here for the Back to Basics Archive ]

Infrared Detector Resolution: How Important Is
It for Accurate Thermal Measurement?

by Jon Chynoweth*

 

When reviewing this article, two comments came quickly to mind. The author has presented his materials in a concise and easy to understand format. Second, how technology in thermal imaging as well as other methods in our industry continues to expand. Test personnel have more detailed data available allowing extremely accurate condition determinations not possible a few years ago. Roderic K. Stanley Associate Technical Editor

Figure 1

Introduction

Until recently, most buyers of infrared imaging cameras had to make a budgeting choice between a low resolution 160 x 120 pixel camera or no camera at all, but they soon learned that they missed a lot at low resolution. Starting at about $25,000 and going up from there, professional, high-resolution 320 x 240 pixel cameras were too costly for anyone who was not a full time thermographer or not working for a company that could afford to buy expensive tools that might be underutilized. That has begun to change, with the introduction of cameras costing less than $15,000 that use a high performance 320 x 240 microbolometer detector. Still, many users want to know why higher resolution is so important in thermal imaging. Why have the professional thermographers always preferred the 320 x 240 pixel detectors used in uncooled, focal plane array cameras?

---

With infrared cameras, resolution affects temperature measurement accuracy, not just image quality.

---

Pixels and Resolution
Pixels are the data acquisition points for thermal measurement, and these data are used to create a visual image from the thermal profile. Having more data points means more information is provided for accurate thermal interpretation. Having more pixels also results in greater visual resolution in the thermal image - so for a given field of view, smaller details can be identified in the thermal image and accurately measured for temperature.

To make infrared cameras affordable to a wider range of users, detectors with pixel counts of 160 x 120, with or without interpolation, were introduced as a low cost alternative. But this is one-fourth the resolution of the 320 x 240 detector: 76 800 pixels versus 19 200. The larger detector produces an image twice as wide and twice as high, with four times the data for a given field of view.

 

Advantages of High Resolution Cameras
For the working thermographer, high resolution allows a camera to work at a much greater distance from a target without a loss of temperature measurement accuracy. A target must cover at least 9 pixels on the focal plane array to be accurately resolved, or the resulting temperature measurement will be compromised when the camera averages in extraneous background. The lower resolution detector interpolates a greater area between pixels and averages in temperature readings unrelated to the target. In practice, a target that's 160 mm² (0.25 in.²) can be accurately measured for temperature at a distance of 18 m (60 ft) with the 320 x 240 detector, while the 160 x 120 detector has to be at 9 m (30 ft) to ensure the same accuracy. (These distances are based on a 25 by 19 degree lens.)

Everyone knows how photographers boast about the resolution of their newest digital cameras; resolution is even more important in infrared imaging. With infrared cameras, resolution affects temperature measurement accuracy, not just image quality. The 320 x 240 detector, with 76 800 temperature measuring pixels, resolves an area smaller than 65 mm2 (0.1 in.²) at 1.8 m (6 ft), while the 160 x 120 detector, with just 19 200 pixels, can't image anything smaller than twice that size. With more background averaged into temperature readings, the readings are inherently less accurate (Figure 1). Likewise, the low resolution thermal image looks as though it's made of tiny little squares, no matter what the viewing size. See Table 1 for more details.

Table 1 Resolution as related to target distance, field of view and pixel size
Target distance (m [ft])	Field of View (m [ft])	Pixel Size 320 x 240 Arrangement (mm [in.])	Pixel Size 160 x 120 Arrangement (mm [in.])
0.3 (1)	0.1 x 0.09 (0.38 x 0.29)	0.35 x 0.35 (0.014 x 0.0.14)	0.74 x 0.74 (0.029 x 0.29)
1.8 (6)	0.7 x 0.53 (2.3 x 1.73)	2.18 x 2.18 (0.086 x 0.086)	4.39 x 4.39 (0.173 x 0.173)
3.0 (10)	1.2 x 0.88 (3.83 x 2.88)	3.66 x 3.66 (0.144 x 0.144)	7.32 x 7.32 (0.288 x 0.288)
6.1 (20)	2.3 x 1.76 (7.67 x 5.76)	7.32 x 7.32 (0.288 x 0.288)	14.61 x 14.61 (0.575 x 0.575)
15.2 (50)	5.8 x 4.39 (19.17 x 14.41)	18.26 x 18.29 (0.719 x 0.720)	36.50 x 36.60 (1.437 x 1.441)

Lastly, another fine point in detector design is fill factor - the space between pixels on the detector that has to be interpolated when the image is formed electronically. The newest detectors have reduced this spacing by 25%, giving them the highest fill factor and best imaging capability in the industry. All of these factors result in the latest generation of infrared cameras being both more affordable and more powerful for use in NDT.

 

* Mikron Infrared, Inc., 1101 Elevation St., Hancock, MI 49930; (888) 506-3900; fax (906) 487-6066; e-mail <jon@mikroninfrared.com>.

 

Copyright © 2006 by the American Society for Nondestructive Testing, Inc. All rights reserved.