Unmanned Systems Technology 020 | Alpha 800 I Additive Manufacturing focus I USVs insight I Pegasus GE70 I GuardBot I AUVSI Xponential 2018 show report I Solar Power focus I CUAV Expo Europe 2018 show report

8 Platform one Measuring the temperature of a surface with a UAV increases the efficiency, safety and productivity of the task (writes Nick Flaherty). It can however be challenging to take measurements of a highly reflective surface because the image is influenced by background thermal reflections. In a UAV application, an unpainted and clean metal roof for example can appear colder than it actually is because the shiny roof reflects the sky above it. With a stainless steel sheet on a rooftop with 0.80 reflectivity and 0.20 emissivity, a radiometric temperature measurement would be strongly biased towards the reflected background temperature of the sky. A clear sky might have a background temperature that is typically below 0 C, and possibly as low as -20 C. The actual sky background temperature also has an impact. Radiometric temperature measurements of glints from the sun can be out by hundreds of degrees. The thermal image is also affected by fluctuations in infrared radiation caused by the air’s density and its relative humidity, as well as the distance between the object’s surface and the camera. The transmission of infrared radiation between the surface and the camera can change the radiometric temperature measurement, and is a measure of the effective heat reaching the camera On a warm and humid day, for example, a 100 m air path at 35 C and a humidity of 80% has a theoretical transmission of 80%. That means only 80% of the thermal radiative heat emitted from a surface will reach the camera. If the atmospheric transmission loss is not accounted for, a UAS looking at a 50 C object with a known emissivity of 0.97 will read 47.6 C – a 2.4 C error caused by the air path alone. The best way to mitigate the atmospheric transmission effect is to minimise the distance between the UAV’s camera and the surface. For example, at 10 m the transmission is 96% and the radiometric temperature uncorrected for the air path in the example above will be 49.5 C. The atmosphere can affect the temperature measurements in other unexpected ways. Rain, snow, smoke, dust or any other particles in the air can reduce the atmospheric transmission and change the background temperature. Laboratory measurements by thermal imaging specialist FLIR on UAV cameras suggest that a measurement spot in the thermal image should be at least ten pixels in diameter to report a meaningful measurement. FLIR tested a Vue Pro R camera with a 13 mm lens, a 640 x 512 pixel resolution sensor and a 17 µm pixel pitch, at an altitude of 20 m. It found that a 30 cm 2 surface directly below the camera will only be 12 x 12 pixels in a thermal image. The focus and blur of a thermal image can also increase the number of pixels needed to make accurate radiometric measurements. For example, the shutter speed of the Vue Pro camera is 1/30 s, so a fast-moving UAV could record a blurred image with reduced accuracy. A hot surface that is smeared out because of the UAV’s motion will look cooler, while a cool surface may appear hotter. FLIR concluded that a sufficient thermal spot size, and gimbal mounts and active stabilisation techniques to reduce camera jitter, allow accurate temperature measurements to be taken from a UAV. Beating the surface heat Airborne vehicles June/July 2018 | Unmanned Systems Technology There are various challenges with measuring surface temperatures using a UAV, according to FLIR