Thermal Imaging is the conversion of radiated or reflected heat into real-time pictures or images. A thermal image is an analogue pictorial representation or visualization of temperature differences.

All objects above absolute zero (-273 degrees) emit radiation, some of which is infra-red. Depending on temperature and emissivity, most objects in the world can be thermally imaged.

Infrared Spectrum

Infrared covers four regions of the spectrum; (a) Near infrared, 0.7 -1 micron, this is nearest to the visible wavelength. (b) Short wave, 1.0 - 2.5 microns. Both of these wavelengths regions rely on reflected solar radiation and can only be used in daylight or illuminated conditions. (c) the Medium or short wave thermal infrared, 3 to 5 microns, detects radiation emitted from objects and can be used in total darkness or daylight. This wavelength is often used for high temperatures such as boilers, kilns etc. (d) the Long wave thermal infrared, 8 -14 microns, is most commonly used in industry since the detectors are efficient at environmental temperatures and can also be used for high temperature operations with appropriate filtering.

Thermal Imaging Optics

Most materials are opaque to medium and long wave infrared including glass and water. Optical materials such as germanium and some other exotic materials such as zinc sulphide, zinc selenide, magnesium fluoride and sapphire are used since they are mostly transparent to the thermal wavelength. These materials are very expensive. Some low grade commercial thermal imagers utilize composite materials to lower the costs but there is no compromise where quality is required. Most military specification thermal imagers use coated germanium and optical magnification rather than digital magnification.

Industrial Thermal Imagers

Over the past few years there has been an incredible advancement in thermal imaging technology. Miniature electronics has allowed for cameras to become very much smaller and far more efficient however the optics and detectors are still very expensive.

Industrial applications can be broken down into two basic categories; Ground and Aerial. Although it is not quite this simple, most ground applications can be accommodated using good quality hand-held equipment. Aerial surveys however require high thermal and spatial resolution to provide quality data at long range.

The most common types of thermal imaging equipment used in industry are: Pyro-Electric Vidicon, (PEV), Focal Plane Array (FPA) and the SPRITE (Signal Processing In The Element). The PEV is the lower end of the price scale and image quality but is perfectly acceptable for Electrical Condition Monitoring and close proximity surveys. The FPA is usually of mid range and is used for general survey applications. The SRITE detector is generally used in top grade military equipment. Each type has advantages and disadvantages and more detail can be given on request.

Interpretation of Thermal Images

Aerial or Ground infrared, also known as Thermal Imaging or Thermography, is best suited to give qualitative rather than quantitative data. Infrared non-contact quantitative systems need accurate information of surface emmissivities if radiant energy is to accurately relate to surface temperature.

Thermal images recorded in 8 -14 micron wavelength (Long Wave) have no visible content and no natural colour. Because of the long wavelength, thermal or infrared images cannot compare in spatial resolution to visible photographs that are recorded in 0.4 -0.8 micron band. Thermal resolution and spatial resolution are inter-dependant to produce good quality thermal images, therefore if a house roof is close to or the same temperature as the nearby road, for example, the image will appear dull grey with little detail. Temperature contrast will produce picture or image contrast and high detail.

Thermal imagers detect and record Infra-red radiation emitted from the surface of any subject being viewed. The imager does not have the ability to see below the surface. However, the radiation from the surface is often influenced by sub-surface detail, which effects the thermal characteristics of adjoining material(s).

When looking at a large area, the emissivity of various surfaces must be considered. Most materials found on the surface of buildings will have a relatively high emissivity but there will still be noticeable differences in the perceived image due to a change in surface material. This can be overcome by a detailed knowledge of the building under investigation. Some metals and glass can reflect infra-red radiation and apparent 'hot spots' can be a reflection from a hot object nearby.

Infra-red aerial surveys provide a 'global' visualization of heat radiation from building surfaces. It is useful data for determining areas of concern or for determining work priority. We can also provide detailed ground level thermal imaging surveys in support of aerial data.

Infra-red surveys of heated buildings are always conducted during the evenings of the Winter/Autumn months of the year.

Aerial thermal Imaging for surveying buildings for heat loss and moisture operate in 8-14 micron wavelength and detect heat only, visible light is not detected at all. Heat energy from daytime sunshine can be absorbed in brickwork and therefore a survey is conducted well after sunset to ensure that all effects of solar energy have dissipated. 

Infra-red - just like visible light - absorbs, reflects and re-radiates from materials in amounts depending upon their colour and structure. Brown building bricks for example absorb and retain heat energy more readily than lighter colored building materials, this must be considered when analyzing data since they may effectively appear at different temperatures simply due to their emissivity. Water bodies such as rivers or lakes retain heat and are slow to change with ambient temperature changes, whereas ground surface temperatures can change rapidly. This is why water often appears warmer than its surroundings.

A thermal Imager detects and displays surface temperatures only. The surface temperature under normal conditions is the result of heat energy conduction through the walls from a heated internal room. Moisture is an excellent conductor of heat and when insulation is damp it can become a conductor rather than an insulator. 


When analyzing thermal data, monochrome images are normally preferred because of the wide range of grey tones of temperature that can be differentiated by computer. There are of course no natural colors in the long wave infra-red wavelength so we apply a palette of colors to the grey tones. Colour images are very much easier for the naked eye to interpret so both color and monochrome are supplied.

To assist in analyzing your aerial data a temperature palette is provided for assessing temperature differences.

Emissivity

A measure of the ability of a surface to radiate energy as measured by the ratio of the radiant flux per unit area to that radiated by a black body at the same temperature. 

Example:- Black body = 1.00
Red Rough House Brick = 0.93
Polished Aluminum = 0.095
 

Emissivity is a measure of the thermal emittance of a surface.  It is defined as the fraction of energy being emitted relative to that emitted by a thermally black surface (a black body).   A black body is a material that is a perfect emitter of heat energy in that it emits all energy it absorbs and has an emissivity value of 1.  In contrast a material with an emissivity value of 0 would be considered a perfect thermal mirror and imaging this material would result in readings of reflected energy only and not the actual material.  For example, if an object had the potential to emit 100 units of energy but only emits 90 units in the real world.  That object would have an emissivity value of 0.90.  In the real world there are no perfect "black bodies" and very few perfect infrared mirrors so most objects have an emissivity between 0 and 1.

As stated before it is hard to determine the emissivity of all objects in your field of view.   It is also difficult to determine the emissivity of a single known material.   This is because emissivity is a measure of the "surface" emittance of an object.  The surface of objects (especially metals) changes with the passing of time.   For example, if you look at the table below you will see that corroded copper (E 0.78) has a significantly different emissivity value than shiny copper (E0.02).  This difference introduces a judgment call on the part of the thermographer.  How do you determine how corroded or shiny a piece of copper is? 

The answer is you can come close but you cannot be exact.  Additionally, the numbers in the emissivity table are approximates or averages, in the real world values may be skewed.  So how do we manage emissivity.  It is possible to determine actual emissivity values for a material but it is not practical in the real world.  The process requires a spectrometer (expensive and usually not available) or dismantling the object and testing each piece.  Even if you determine an accurate E value, the next time you scan the item that value will have changed rendering your old value useless.  In most infrared applications exact temperature measurement is not necessary. 

For example, if a circuit has a fault limit of 150° F and your instrument measures 100° F and the E value skews the temperature reading by 5°F  you are left with a +- 5°F variance which in this case is negligible.  Additionally, most thermal infrared applications rely on temperature difference (Delta T ) rather than exact temperature readings.  To use our previous example of the circuit we measured, there would most likely be more than one circuit next to each other.  If you use the same E value for both circuits they will both be skewed the same amount.  If the one circuit was reading 100°F (which we will assume is normal operating temperature) and the adjacent circuit reads 150°F we are left with a Delta T of 50°F which would indicate a problem and as you can see negates the emissivity problem.  E values become even less of a problem when trending an area over time.  If the same circuit with the reading of 100°F has a reading of 110°F the next time you scan it and a reading of 115°F the next time, with the same emissivity setting, we know a problem is developing regardless of the error introduced by emissivity.  

Dealing with emissivity is not as hard as it would seem.  The important things to remember are that exact temperature measurements are difficult to obtain (do not promise this to your clients unless you are sure you can back it up), temperature difference (Delta T) is more important than exact readings in most applications, and that trending an object can reveal problems regardless of E value error.  In the real world you pick an emissivity value that approximates the scene you are imaging and then you record it and maintain that same setting every time you scan that object.

Seeing Heat Energy