Today, thermal cameras are widely used in oil and gas, marine, utilities, high security and critical infrastructure settings to support everything from process safety to offshore threat detection. But what exactly are thermal cameras? Why are they popular? And, what are the factors organizations need to consider when deciding whether or not to deploy them?
“Their strength lies in proactive surveillance ‒ as an early warning mechanism for approaching objects, mechanical failure and process issues.”Wayne Barraclough, Product Manager, Synectics
Like many security and surveillance technologies, thermal imaging was first introduced through military applications, where negating the need for light in order to identify potential threats had clear operational benefit for field maneuvers.
What might be surprising, however, is that the very first thermal camera was actually developed almost 60 years ago. And while the technology has stood the test of time, a lot has changed.
We talk to Product Manager Wayne Barraclough, and Systems Consultant John Tinson to find out about the latest developments in thermal camera technology.
Some people use thermal cameras and Infrared (IR) cameras interchangeably – are they correct in doing so?
WB: Not really, but there is a good reason for the confusion. Thermal cameras do use radiation from the infrared spectrum. In that sense they are ‘technically’ IR cameras. But the range we are talking about is the very far end of that spectrum – well beyond the frequency range of ‘near infrared’ i.e. the area of the spectrum where most IR cameras operate.
The IR range is actually much closer to visible light on the spectrum. As a rough guide, visible light (what the human eye can easily see) spans from about 450nm to 750nm, with near infrared kicking in at approximately 700nm and going up to around 1000nm. By comparison, thermal cameras operate at wavelengths up to 14,000nm.
Also let me just go back to the word ‘radiation’ as this is important. Thermal cameras operate by detecting differences in the radiation emitted by an object; IR cameras detect light (near-infrared light) reflected from an object – in complete darkness this light has to be generated by a light source, typically an integral infrared LED.
Does this difference impact on the camera design?
WB: Yes, significantly. An IR camera is built and functions the same way as a traditional electro-optical camera, but will have the IR filter removed. Thermal cameras require an entirely different design – most notably using a material such as germanium instead of glass for the optics in order to accommodate the operational wavelength. The sensors used also need to be different in order to respond appropriately to photon detection. This is one of the reasons why thermal cameras are more expensive than standard electro-optical cameras.
If thermal cameras are more expensive than IR LED cameras which can also operate in complete darkness, why use thermal imaging?
JT: It depends on the application but in terms of threat detection – distance is probably the single biggest reason. Because they operate by measuring emissivity rather than capturing reflected light, thermal cameras can detect objects at much greater distances.
Active illumination with cameras fitted with IR LED illumination only have a limited operational range typically out to about 200m and very much depend on the reflectance of the target. As the distance increases and the camera field of view is reduced or zoomed in, the light attenuation increases and a more powerful illuminator is required, so it becomes a law of diminishing returns. On the other hand, thermal cameras passively operate in the MWIR and LWIR bands with their operational range only limited by the lens and sensitivity of the camera.
WB: It’s also important to mention LED degradation. In as little as 5 years, LED light sources will typically degrade by approximately 30% which can significantly impact on camera capability and image quality. Thermal cameras will require a more significant initial outlay but their lifespan is much longer.
Are there different types of thermal camera?
JT: There are two types of true thermal cameras – cooled and uncooled.
Cooled thermal cameras offer a better signal to noise which translates into a higher sensitivity. To achieve this, the imager is cooled to about 70K using a Stirling engine (cryogenic-cooler) which reduces the camera’s operating temperature to ensure the sensor isn’t affected by the camera’s own radiation. Most cooled cameras offer around 3-4 times the sensitivity of uncooled thermal cameras. This increase in sensitivity enables longer focal length lens to be used over much greater target distances.
So cooled thermal cameras should typically be used for identification at distance?
JT: Yes, for long distance applications cooled thermal cameras are often the best choice but we must be careful of the term ‘identification’. Cooled thermal cameras offer greater thermal sensitivity for the same number of pixels i.e. they can detect much smaller differences in radiation/temperature at greater distance. This facilitates improved threat detection but not necessarily identification.
Organizations using electro-optical cameras will be familiar with DRI ranges – Detection, Recognition and Identification criteria in terms of percentage height of the object. Here ‘Detect’ means with high probability a person is present, ‘Recognize’ enables the observer to confirm the person is someone they have seen before and ‘Identify’ enables the observer to confirm the identity of a person beyond reasonable doubt.
With thermal cameras, the DRI is based on the Johnson Criteria where the resolution is defined by a minimum line pair (lp) or pixel value where a line pair is defined as one white line adjacent to a black line on the thermal image captured. While the minimum lp value for ‘Detection’ is 1, for ‘Recognition’ it is 3, and ‘Identification’ requires 6. Here the DRI means something different to above where ‘Detect’ confirms that an object is present, ‘Recognition’ enables the type of object or class to be confirmed e.g. man or car and ‘Identify’ enables a specific object or class to be discerned e.g. type of car.
As I mentioned earlier, a cooled thermal camera offers greater thermal sensitivity for the same number of pixels but in order to recognize or identify an object, other factors have to be considered that the Johnson Criteria doesn’t account for – from technical aspects such as spatial resolution, field of view and spectral detection range, to so-called ‘conditions of use’ such as weather, backdrop and ambient temperature.
WB: In reality, fulfilment of the DRI process from a systems perspective is typically achieved using a combination of camera types. For instance, as part of a perimeter protection solution a cooled thermal camera may be used to detect long-range threats and trigger initial security alerts, with the ‘recognition’ and ‘identification’ function automatically switched to electro-optical cameras if the object detected continues to approach the site/asset.
In what kind of scenarios might an uncooled thermal camera be more beneficial?
JT: There are many vital process and maintenance applications where distance and superior sensitivity are not key factors, but where the overall benefits of thermal camera technology are still required. For example, flare stack analysis in the oil and gas sector, or monitoring the liquid levels in tanks.
In both of these scenarios operating capability in poor/no-light conditions remains essential, and the objective itself may be non-visible detection i.e. heat levels, gas compositions (different gases will have different emissivity), but the detection range is shorter and precision imaging is not the priority. Uncooled thermal cameras provide an ideal solution, and a more cost effective option.
There are no hard and fast rules – it’s all about weighing up objectives, operating conditions and budget.
Are there any scenarios where thermal cameras shouldn’t be used?
WB: As John says, once organizations understand the capabilities and limitations of thermal imaging, it comes down to what organizations want to achieve.
That said, while thermal cameras do offer many advantages, many of those are lost when looking at surveillance for retrospective use only i.e. for recording and review. Thermal cameras won’t, for example, offer the same detailed evidentiary review capabilities as electro-optical cameras.
Their strength lies in proactive surveillance ‒ as an early warning mechanism for approaching objects, mechanical failure and process issues. Thermal cameras are an ideal first line of defence in terms of detection capability and in that sense perhaps things haven’t moved that far from the camera’s military origins after all.