Thermal Cameras Explained

This blog is a response to an email asking how thermal cameras work.

A device that intercepts and interprets thermal radiation is called a radiometer, a thermal imaging camera is in effect a multipoint imaging radiometer. The basic principle is that it converts invisible thermal radiation into a visible image. Most thermal imaging cameras work in either the midwave (2-5 micrometre) or long wave (8-14 micrometer) infrared region. This is the part of the electromagnetic spectrum most associated with heat. There are two basic families of detectors used in thermal imaging, thermal detectors and photon detectors. Thermal imaging uses limited spectral bands, and works differently than night vision systems. Night vision systems usually operate in the near infrared waveband. They intensify light in this region, and some light is required. Thermal imaging needs no visible light to work and will work in total darkness. Other devices that work in the near infrared are TV remote controls, although the light emitted by a remote control in not visible with the human eye, it can often be seen with a camera phone. This happens because the CCD used in digital cameras is in fact an image converter. It’s spectral characteristics are different than the eye and in order to get the colours looking right they work slightly outside the region where the eye works, and into the near infrared region.

Thermal detectors are more common today due to manufacturing costs. The detector is made up of a number of pixels. The optic focuses the infrared onto the pixels, each pixel heats up in response to the energy focused on it. As the temperature of the pixel increases its electrical resistance also increases, this is monitored by advanced signal processing and an image is produced. An increase in the temperature of the object of 1 degreeC will cause the pixel to increase by 0.01degreeC. The advantage with these type of detectors is that don’t really need cooling, other than perhaps a peltier. The disadvantage is that they can be a bit slow, as it takes time for the pixels to heat up.

A photon detector works in a totally different way, it intercepts the photons being emitted by objects and converts these directly to an electrical signal. The incoming photons cause a specific quantum event, such as electron interband transitions. Normally there are three layers, a conduction band, an energy gap, and a valence band. If the incoming photon has enough energy to jump the energy gap, it will knock an electron into the conduction band, generating a signal. The advantage of this type of detector is that it is very sensitive, and can operate at high speeds. The disadvantage is that they usually require cooling to cryogenic temperatures to remove parasitic noise. Cooling systems used are; Stirling coolers, joules Thomson cryostats and liquid nitrogen.