| |
|
| |
Renee Visser handles the
thermal imaging camera.
|
Dingoes make reluctant movie stars. For starters, they are extremely wary of humans, sometimes avoiding an area where people have been for several days afterwards. To further confound matters, they are also most active at night, which presents all sorts of challenges for the observer.
Ecologist Renee Visser from the School of Resources, Environment and Society contemplated these stumbling blocks at the start of her PhD project, which is investigating how wild dog populations regulate smaller predators, like feral cats and foxes. Other studies have shown dingo culling can lead to increases in the number of introduced predators, and a decline in the numbers of their smaller native counterparts. Visser is investigating the role that dingo social structures play in this complex ecological web. But how to observe such a notoriously shy creature?
“Traditionally, there are a number of options,” she says. “You can get night vision binoculars and stay awake all night, or you can get a video camera. But with conventional cameras, you have to set-up motion detectors, infrared lighting, and you’re often limited in that you have to sift through all of this video footage.”
“Quite often the motion sensor only works within a relatively short field, and they can pick up unwanted things like trees blowing in the wind. It’s also very time consuming. There are a lot of different components to set-up.”
The heat of the matter
If the project was to succeed at all, she needed a way to record dingo behaviour without the deterrent of all-night vigils or cumbersome and potentially inefficient recording equipment. Serendipity intervened in the form of an ARC linkage grant, a thermal imaging camera, and an intrepid Masters student called Tim Raupach, who is based at National ICT Australia (NICTA).
|
|
|
Tim Raupach as
seen by the camera.
|
|
Together, the ecologist and computer scientist are pioneering the groundbreaking use of thermal imagery to record animal behaviour. Making use of the cutting-edge video technology, Raupach has refined motion-detection software that will only start recording data when animals of a certain size and temperature range enter the camera’s field, lured there by a bait carcass. This does away with the need for an external motion detector. It appears the thermal approach has several other advantages, too.
“It picks up purely heat,” Raupach says, “so it can see in the dark, and doesn’t need any external illumination. You can also get the exact temperature of every pixel in the image – it will actually tell you the surface temperature of everything that comes back. Nobody has ever done any work with temperature and dingoes before. So even though initially we’re using it to capture their motion and look at how they behave, there is also some opportunity to look at their temperature as well.”
Visser believes information about where heat is distributed on the body of a dingo could hold some interesting clues to their behaviour, linking to traits like aggression and stress. But she says it is too early to predict whether this line of inquiry will prove fruitful. For now, she is extremely pleased to have the video recording working smoothly after several months of testing
Setting the scene
“Thermal has been used in medical imaging, and in ecology for aerial counts, but very few behavioural studies have been conducted,” Visser says. “This is the first time this has been used for monitoring behaviour of a cryptic species: those that are really hard to sit and watch because they take a long time to get used to people.”
Initial field tests have taken place in the Bago State Forest in New South Wales, where the camera needed to be safeguarded in the hilly, rain-soaked terrain. The team developed a simple marine-ply casing to protect their equipment. For Raupach, the chance to get out in the wild was a refreshing challenge.
“I had to see what we were dealing with first hand,” he says. “Working out how to protect everything from the elements was a big part of what we’re doing. Often computer science can be very theoretical, and this is really practical. One of NICTA’s goals is to do application-based research. As for the setting it up, it was raining, and it was muddy. It’s a far cry from the lab.”
Despite the inhospitable surrounds and the shy nature of the wild dogs, the team were genuinely surprised when results were achieved quickly.
“I think we all cheered that it actually worked,” Visser says. “We’ve used some footage from captive animals to trial the camera and this helped determine how we set the camera up in the field. It then took a long time to make adjustments for working in a remote area as it has to be powered correctly, protected from the elements, and then we had to draw the dingoes to the carcass. We just didn’t know if it was going to work or not. Then it did! We got footage on the first night.”
Lights, camera, action
| |
|
| |
A thermal image of a dingo.
|
The images captured so far by the thermal camera show haunting ghost dogs, which glow brightly against the blackness of their environment. Bright too are the team’s hopes for the next stage of the project, where the camera will be taken to the arid zone in South Australia to record larger dingo populations. Visser says it will be there, where dingoes have minimal contact with humans, that she hopes to gain a fuller understanding of the behaviour and society of these cryptic animals. Until then, she is full of praise for her research partner and optimistic about what thermal imaging could mean for other researchers in the field.
“It’s been worth it. In the end we’ve got something that’s really spot on. It’s working really well, and we’re both happy,” she says.
Raupach agrees enthusiastically with the possibilities hinted at by the project.
“I think that ecology and computer vision are usually seen as two vastly separate fields,” he says. “There could be a lot achieved in bringing them together.”
Back to top
Back to Reporter home |
|