Thermal cameras detect infrared radiation, a form of energy that every object with a temperature above absolute zero naturally emits. They don’t detect “heat” directly. Instead, they capture invisible light in the 8 to 15 micrometer wavelength range and translate it into a visible image where colors or shades represent temperature differences. This means thermal cameras work in complete darkness, through smoke, and in conditions where the human eye sees nothing.
How Thermal Cameras Build an Image
Every object radiates infrared energy proportional to its temperature. A thermal camera’s lens focuses that radiation onto a grid of tiny sensors, each one acting like a pixel. These sensors absorb the incoming infrared energy and convert it into an electrical signal. The camera’s processor then assembles those signals into a temperature map called a thermogram, assigning colors (or grayscale values) to each pixel based on how warm or cool that spot is.
Modern thermal cameras scan this sensor grid 30 times per second, detecting temperatures from roughly negative 20°C up to 2,000°C. The best sensors can resolve temperature differences as small as 0.1°C, which is why thermal imaging picks up subtle variations invisible to the naked eye, like a fingerprint left on a table or a patch of damp drywall behind paint.
What They Can and Cannot See Through
One of the most common misconceptions is that thermal cameras can see through walls, glass, or water. They cannot. Standard glass blocks infrared radiation in the 8 to 14 micrometer range, so pointing a thermal camera at a window shows the temperature of the glass itself, not what’s behind it. The same applies to thick water layers, solid walls, and most plastics. What thermal cameras can see through is smoke, dust, and light fog, because these particles are small enough that long-wave infrared passes around them.
They also cannot see through clothing to detect skin temperature underneath, though they will pick up heat radiating through thin fabric. And they don’t “see” colors, textures, or written text. The image is purely a temperature map.
Emissivity: Why Some Materials Are Harder to Read
Not all surfaces radiate infrared energy equally, even at the same temperature. A property called emissivity determines how efficiently a surface emits infrared radiation on a scale from 0 to 1. Human skin has very high emissivity (around 0.98), which makes people easy to detect and measure accurately. Polished aluminum, on the other hand, has an emissivity of just 0.05. It reflects infrared from surrounding objects rather than emitting its own, making it appear misleadingly cool on a thermal image.
This is why shiny metal surfaces, mirrors, and reflective insulation can fool thermal cameras. Professionals using thermal imaging for inspections typically adjust the camera’s emissivity setting for the material they’re scanning to get accurate readings.
Thermal Crossover: When Detection Fails
Thermal cameras rely on temperature contrast between an object and its surroundings. When both reach the same temperature, the object effectively disappears from the image. This phenomenon, called thermal crossover, typically happens twice a day, most often around sunrise and sunset, when the environment is warming up or cooling down and objects briefly match their background temperature.
Metal objects with low thermal inertia are especially prone to this during sunny conditions because they heat and cool faster than their surroundings. Rain, fog, and rapidly changing weather can trigger crossover at unpredictable times throughout the day. For anyone relying on thermal imaging for surveillance or search and rescue, thermal crossover represents a real blind spot.
How Thermal Differs From Night Vision
Thermal imaging and night vision goggles solve the same problem (seeing in the dark) through completely different methods. Night vision devices amplify existing light. They collect tiny amounts of visible and near-infrared light bouncing off a scene, then multiply those photons roughly 1,000 times through an internal tube, producing the familiar green-tinted image. They need at least some ambient light, whether from stars, the moon, or streetlights, and they produce an image that looks like a dim version of what your eyes would see.
Thermal cameras need zero light. They detect emitted radiation, not reflected light, so they work in absolute darkness. Their output is a false-color map of heat rather than a recognizable visual scene. A person hiding in bushes at midnight is invisible to the naked eye and difficult for night vision, but stands out immediately on thermal because the human body is significantly warmer than vegetation. The tradeoff is that thermal images lack visual detail. You can tell a person is there, but you won’t read their face or see what they’re wearing.
Fever and Medical Screening
During disease outbreaks, thermal cameras became a fixture at airports and building entrances for fever screening. The CDC considers 100.4°F (38°C) the threshold for fever in most work settings, and thermal cameras can flag individuals whose facial temperature exceeds that cutoff. The approach works because skin has such high emissivity, making the readings reliable when the camera is properly calibrated.
Accuracy depends heavily on setup. The camera needs a controlled environment (not outdoors in wind), a consistent distance from the subject, and ideally a reference temperature source nearby for calibration. Thermal cameras measure skin surface temperature, not core body temperature, so readings skew lower. Someone who just came in from cold weather may not register as feverish even with an elevated core temperature. These systems work best as a first-pass filter rather than a diagnostic tool.
Building Inspections and Heat Loss
Thermal imaging is one of the most practical tools for finding where a building loses energy. Poorly insulated walls, air leaks around windows, and gaps in weatherstripping all show up as temperature anomalies on a thermal scan. In one study of residential buildings, more than 87% of wall surface area showed temperature differences exceeding 2°C from the baseline, revealing widespread insulation deficiencies that were invisible during a visual inspection.
Windows and wall junctions are the most common trouble spots. A thermal camera can also detect moisture inside walls, because wet insulation conducts heat differently than dry material, creating a visible cool patch on the exterior surface. For homeowners, a thermal scan during cold weather (when the temperature difference between indoors and outdoors is greatest) reveals the most useful information.
Electrical and Industrial Faults
Loose connections, overloaded circuits, and failing components in electrical panels generate excess heat before they fail visibly. Thermal cameras detect these hot spots during routine inspections, often catching problems months before they’d cause a breakdown or fire. In electrical work, a temperature rise of 9°C above ambient is generally considered normal, while a rise of 20 to 27°C signals an emerging problem that needs attention.
Industrial facilities use thermal imaging to monitor motors, bearings, steam traps, and pipelines. A bearing running hotter than its neighbors suggests friction from wear. A steam trap showing no temperature differential across its inlet and outlet has likely failed. These scans take seconds and can be done while equipment is running, avoiding costly shutdowns for manual inspection.
Other Common Uses
- Search and rescue: Finding people in darkness, rubble, or dense forest based on their body heat, even when they’re not moving or calling for help.
- Wildlife monitoring: Counting animals at night without disturbing them, since warm-blooded creatures contrast sharply against cooler terrain.
- Firefighting: Seeing through smoke to locate the seat of a fire, find trapped occupants, or identify hot spots in walls that could reignite.
- Law enforcement: Tracking suspects at night, locating recently driven vehicles by their warm engines, or finding buried evidence that retains heat differently than surrounding soil.
- Agriculture: Detecting irrigation problems, plant stress, and pest damage across large fields, since affected plants change temperature before they show visible symptoms.