Technology now frequently used in wildlife conservation is evolving at a rapid pace, with improvements promised to make research applications and integration with other technology easier, faster, and smarter. In this case study, World Wildlife Fund’s Megan Ossmann shares about her work with colleague Eric Becker testing out FLIR’s new Duo Pro R thermal camera as a detection tool for loggerhead sea turtle nests.
In recent years, researchers have increasingly taken advantage of new technology for wildlife detection and censusing - particularly UAVs and thermal cameras. This past April, FLIR Systems, a leader in the thermal-imaging world, released a camera that combines both thermal tech and UAV applications in one innovative package: the FLIR Duo Pro R. Designed to be mounted on a drone, the Duo Pro R combines a high resolution, radiometric thermal imager, 4K color camera, and full suite of onboard sensors. The camera is capable of capturing both visible and thermal data in a single flight, as well as automatically geo-tagging each captured image.
In July, I flew down to the Florida panhandle to join Eric Becker, WWF’s in-house Conservation Engineer, in testing out the Duo Pro R. Being surrounded by marine life, we decided to test whether it was possible for the thermal camera to detect sea turtle nests on the beach. This idea stemmed from literature stating that the area containing a nest will typically be 1-2o F warmer than the surrounding sand, particularly at the late stages of incubation (Matsuzawa et al. 2002, Sandoval et al. 2011).
Round 1: On the Ground
We worked with the Navarre Beach Sea Turtle Conservation Center, whose founder, Cathy Holmes, showed us the location of two neighboring loggerhead sea turtle nests a couple of weeks from predicted hatching. Right after sunrise, we hauled our equipment (which included the camera, a battery, wires, a tripod, and a monitor) down the beach toward the nests. Since Eric did not yet have a way to attach the camera to his drone, we set the camera up on a tripod about 20 ft away and connected it to the monitor to ensure we were capturing images of both nests. When connected to the camera’s Bluetooth, the FLIR UAS app allows you to configure the settings, such as choosing between thermal, visible, and picture-in-picture display modes, changing the IR color palette, and switching between video and single/multiple still image recording modes.
Later that evening, Eric created an attachment for his DJI Phantom 2 that would allow the camera to be mounted with relative security (there’s nothing more nerve-wracking than the possibility of a brand-new $5000+ camera falling off of the drone and straight into the ocean). Using his 3D printer, he printed extension feet for the landing gear and a mounting bracket for the drone. While mounting the camera to the drone, he routed the power and video cables to the drone electronics.
Round 2: From the Air
With a goal of arriving at the beach before sunrise, but getting a late start due to last-minute adjustments to the drone attachment, we again set out to survey the nests. Staying in the nearby parking lot this time, Eric manually flew his drone over and around the nests as a photo was taken every second. Some light wind made this flight a bit tricky, so after about ten minutes Eric brought the drone back down. Not to worry - we collected more than enough still aerial photos of the nests.
These photos are captured as radiometric JPEG (RJPEG) images, which have temperature data embedded in each pixel. To extract this data, we uploaded the RJPEGs to FLIR’s free FLIR Tools software. Starting with the ground-level photos, we manually chose points on the photo both inside and outside of the nest boundary and got a list of temperature readings.
These images showed a ~2° F temperature difference between the nest and surrounding area, supporting the literature values (images above). However, the aerial images we collected showed no clear indication of a temperature difference (image below), leaving us questioning why that was. At this point, my time in Florida had come to an end and I returned to Washington DC to wrap up the last few weeks of my internship.
Round 3: Follow-up
Fast forward two weeks and I have received more imagery from Eric, who had conducted another flight after I left. In these aerial images, a clear temperature difference can be seen between nesting and non-nesting areas when the temperature range of the photo is constrained to a 2.2 oF range. This is much different than the first aerial photo, which had a temperature range of about 6 oF, and had no discernable temperature difference between the nesting and non-nesting areas when constrained to a 2 oF range.
So, what does this all mean?
Well, it’s still inconclusive, really. I want to emphasize that this was in no way, shape, or form a controlled study. Many factors are at play here – time of day, altitude of flight, stage of incubation, nearby vegetation, footprints, and fencing material may all have impacted the temperature readings. However, with better controls and thus more robust findings, future studies could substantially impact the way nests are detected and monitored.
For now, I want to focus on the applications of the FLIR Duo Pro R, recognizing our small pilot project as just the tip of the iceberg in terms of possibilities for this piece of equipment. What sets it apart from other cameras is the onboard sensors that allow easy integration. All you have to do is apply power and attach it to the drone, while other cameras typically require more advanced integration or time-consuming post-processing to obtain the correct 3D orientation and position of the camera for analysis. So, if it was this simple to test out its capabilities of detecting sea turtle nests, how successful might it be in detecting above-ground nests of other species? In population surveys? For added security in areas of wildlife conflict? Whether for research, monitoring, management, or any other application, the FLIR Duo Pro R certainly has the potential to make the process that much simpler.
About the Author
Megan Ossmann is a master's student at Duke University studying ecosystem science and conservation, as well as geospatial analysis. She is interested in the applications of GIS in wildlife conservation and is particularly passionate about combating the poaching and trafficking of wildlife. Her recent position as the Wildlife Technology Intern at World Wildlife Fund has fully launched her into the wildlife tech world, and now that she’s started handling camera traps, thermal cameras, and drones, there’s no way she’s going back.
Are you exploring how drones can be used in wildlife conservation? Join our community to share your work!