How Does Infrared Imaging Work: Plain-Language 2026
People who pick up a thermal camera for the first time usually skip straight to the picture and miss what the picture actually means. The false-color image is the last step in a chain that starts with invisible long-wave infrared light leaving a warm surface and ends with a number stored at every pixel. Understanding the chain — even loosely — changes how the user reads the picture, where they trust it, and where they second-guess it. This guide walks through the process from physics to display, in language that a homeowner or new home inspector can follow without an engineering degree.
Every warm object emits long-wave infrared
Anything above absolute zero radiates electromagnetic energy. Above room temperature, that energy peaks in the long-wave infrared band, roughly 8 to 14 micrometers in wavelength — well past the red end of visible light. A wall at 70 degrees Fahrenheit emits a steady stream of this radiation; so does a person at 98.6 degrees, a hot breaker at 140 degrees, and an ice cube at 32 degrees. The hotter the surface, the more energy it emits, and the shorter the wavelength of its emission peak. This relationship is Planck’s law in a sentence.
Humans cannot see any of this. Our retinas only detect visible wavelengths (roughly 0.4 to 0.7 micrometers). Long-wave infrared travels through the same air, bounces off some surfaces, and is absorbed by others, all without registering to our eyes. A thermal imaging camera sees what humans cannot — the emitted heat signature itself. The home inspection tools overview covers how this physics translates into practical inspection workflows.
The lens focuses heat, not light
Standard camera lenses, made of glass, block long-wave infrared completely. Glass absorbs the LWIR band before it can reach a sensor. Thermal camera lenses are therefore made of germanium, a brittle metallic element that is transparent at 8 to 14 micrometers but expensive. This is one reason thermal cameras cost more than visible-light cameras of equivalent body quality — the lens material alone is dramatically more expensive than optical glass.
The lens focuses incoming LWIR radiation onto a small flat sensor at the back of the camera. The geometry is the same as a regular camera lens: a wider focal length captures a broader scene, a longer focal length zooms into a smaller portion of the scene. Field of view and focus distance work the same way as in visible-light photography.
The microbolometer array does the actual measuring
The sensor itself is a microbolometer — a rectangular grid of tiny detectors, each one a pixel. A 160 by 120 sensor has 19,200 detectors arranged in 120 rows of 160 each. A 320 by 240 sensor has 76,800. A 640 by 480 industrial sensor has 307,200.
Each individual microbolometer detector is a thin film of either vanadium oxide or amorphous silicon suspended on a tiny membrane. When LWIR radiation hits the detector, it warms up by a vanishingly small amount — fractions of a millikelvin. The detector’s electrical resistance changes as it warms. The camera’s electronics measure that resistance change thousands of times per second, convert it to a temperature value, and store the value for each pixel. Repeat 19,200 times for a 160 by 120 image; repeat 307,200 times for 640 by 480. The whole grid refreshes typically 9 or 30 times per second.
This is why thermal cameras with higher pixel counts cost more — each detector is an individually manufactured component on a precise grid, with no shortcuts. The HD thermal imager discussion covers how marketing often blurs the line between sensor resolution and display resolution, which are very different things.
From temperature values to a grayscale picture
Once every pixel holds a temperature value, the camera builds a grayscale image. The hottest pixel in the frame becomes white, the coldest becomes black, and the rest spread across a gray gradient in between. This grayscale image is the raw thermal frame. It is rarely what users actually see, because it is hard to read at a glance.
Most cameras add a “level and span” control here — the user can tighten the gray range around a narrow temperature window (say, 60 to 80 degrees Fahrenheit) to amplify subtle differences, or widen it (40 to 120 degrees) to capture a scene with hot and cold extremes at once. This is one of the most important controls in real inspection work and one of the most underused by new operators. A poorly chosen span makes a damaging anomaly invisible; a well-chosen one reveals it clearly.
False color makes the picture readable
The grayscale image gets mapped to a color palette before display. The familiar “iron” palette runs from black through purple, red, orange, yellow, and white as temperature rises. “Rainbow” palettes run cold-to-warm through blue, green, yellow, red. “Grayscale-with-color-anomaly” palettes leave most of the scene gray and only color the hottest or coldest few percent — useful for finding a single hot spot without distraction.
The palette choice is purely cosmetic from the data’s perspective. Every pixel still holds its underlying temperature value. Palette selection just changes which colors represent which numbers. Inspectors often switch palettes mid-job depending on what they are hunting. Building diagnostics work well in “iron”; electrical hot-spot hunting often uses “high color” palettes that emphasize the top of the range.
Visual fusion overlays make interpretation faster
A thermal image alone shows shape and heat, but it does not tell the user what object they are looking at. A hot rectangle on a wall could be a stud, a vent, or a pipe behind the drywall. To solve this, most cameras above the entry tier include a co-located visible-light camera and a fusion algorithm. The result is an image that overlays the visible-light edges of the scene on top of the thermal data — a “thermal-plus-visible” picture in which the user can see at a glance that the hot rectangle is, say, a wall-mounted light fixture.
FLIR markets this as MSX; Fluke calls it IR-Fusion; other manufacturers use proprietary names. The underlying idea is the same: pair the thermal data with visible context. Most working inspectors strongly prefer cameras that do this well, because it dramatically reduces ambiguity in field reports.
Reading the resulting picture without falling into traps
Three common interpretation errors come up repeatedly in field reports — they are worth knowing because each one looks like a real defect at first glance.
The first is the reflection trap. Polished surfaces (glass, mirrors, glossy paint) reflect infrared from elsewhere in the room. A “hot spot” on a window may be the reflected image of a warm radiator behind the inspector. Moving the camera left and right shifts the reflection, which is how the operator confirms it.
The second is the emissivity trap, covered in DOE guidance for building auditors. Different materials emit infrared at different efficiencies. Bare aluminum reads dramatically colder than its actual surface temperature because it reflects rather than emits. Painted drywall reads accurately because painted surfaces have emissivity near 0.95.
The third is the conduction trap. A cool patch on a wall is not always missing insulation; it might be a stud (wood is denser than insulation and conducts heat more readily, so studs frequently read several degrees cooler than the bays on a cold day). Knowing the construction behind the wall changes the interpretation. The system walkthrough guide covers these traps in more detail.
What the picture is actually showing
It bears repeating because the false-color image is so vivid: a thermal picture shows surface temperature, not temperature inside a wall, not moisture directly, not insulation directly. It shows surface temperature patterns from which a trained operator infers what is happening below. A moisture-soaked wall reads cool because evaporation cools the surface; that is not a direct moisture reading, it is an inference from the cooling pattern. Missing insulation shows up because the wall surface tracks closer to the outside temperature than its neighbors.
This is why infrared imaging is best paired with confirmatory tools. A moisture meter confirms what a thermal pattern suggests. A borescope confirms what an infrared scan implies. The thermal camera is a fast surveyor; the follow-up tools are the verifiers. Both ASHI and InterNACHI standards of practice treat infrared imaging as a supplementary technique rather than a stand-alone diagnostic, which is the right framing.
The role of temperature contrast (delta-T)
Infrared imaging works best when there is a meaningful temperature difference between the interior and exterior of the building, or between an anomaly and its surroundings. A wall scanned on a 65-degree spring day with a 65-degree interior shows almost nothing useful because the surface temperature is uniform everywhere. The same wall scanned on a 20-degree winter day with a 70-degree interior shows missing insulation patterns clearly because the temperature contrast across the wall assembly is 50 degrees.
Building auditors typically aim for at least a 20-degree Fahrenheit difference between inside and outside before performing serious insulation diagnostics. Energy programs run by the U.S. Department of Energy include specific weather-condition requirements for documented audits, and many state weatherization programs follow the same standards. The practical implication is that thermal scans for insulation work in heating climates are best performed in winter, and scans for air leakage work in extreme weather conditions of either heating or cooling type.
For moisture detection the contrast requirement is lower because evaporative cooling produces its own temperature signal independent of indoor-outdoor delta. Moisture scans can work in mild weather, though they work even better when the building shell has a defined thermal gradient.
Frame rate and what it implies for handheld work
Consumer thermal cameras typically refresh nine times per second, which is enough for handheld walking-pace scanning but produces visible lag when the camera moves quickly. Commercial cameras typically refresh at thirty or sixty hertz and produce noticeably smoother video. The nine-hertz limit on consumer units is a regulatory artifact — under U.S. ITAR rules, faster frame rates trigger export restrictions, so manufacturers cap consumer products at nine hertz to keep them broadly exportable.
For handheld inspection work, the difference between nine and thirty hertz matters most when sweeping the camera across a wall. A nine-hertz camera produces noticeable smearing if the operator pans too fast; a thirty-hertz camera handles faster panning gracefully. New operators benefit from moving the camera slowly enough that the nine-hertz refresh keeps up. The thermal imaging camera capabilities guide covers the technique adjustments that consumer-tier cameras require.
How the camera handles emissivity in software
Every infrared measurement depends on the surface’s emissivity — its efficiency at radiating heat compared to a theoretical perfect black-body. Most cameras let the user set an emissivity value (a number between 0 and 1) that the camera applies when converting detected energy to temperature. The factory default is typically 0.95, which works well for painted surfaces, drywall, wood, and skin.
Bare metal surfaces have dramatically lower emissivity (polished aluminum sits around 0.1, weathered galvanized steel around 0.28, oxidized copper around 0.78). Reading these surfaces with the default 0.95 setting produces temperature readings significantly colder than the actual surface temperature, because the camera interprets the lower energy emission as a cooler surface rather than a less efficient emitter. Knowing when to adjust emissivity in the menu — or simply to cover the metal surface with a piece of black electrical tape, which has an emissivity near 0.95 and reads accurately — is a fundamental skill that ASHRAE training programs cover early.
References
- DOE guide to thermographic inspections — U.S. Department of Energy
- InterNACHI infrared certification standards — InterNACHI
- ASHI standards of practice for inspections — American Society of Home Inspectors
- ASHRAE building science publications — American Society of Heating, Refrigerating and Air-Conditioning Engineers