A step closer to dingo recognition

Professor Brian Lovell, one of the world’s foremost experts in covert surveillance techniques using facial recognition software, last year undertook a trial that showed his techniques for searching for persons of interest would work in identifying individual dingoes as well.

Prof Lovell is from the Security and Surveillance Research Group at the School of Information Technology and Electrical Engineering at the University of Queensland, and was last year approached by the Fraser Island Dingo Working Group to undertake the experiment.

“I had seen a study where something similar was done with sheep, and the idea of trying it out on dogs intrigued me,” he said.

More than 300 images of 80 individual dingoes were supplied from the Queensland Parks and Wildlife Service, Griffith University and private research databanks for Prof Lovell to work with, which he was able to successfully match 80 per cent of the time.

“This is about the same as we get with humans – 85 to 95pc.”

Department of Environment and Heritage Protection interest in the technique centres on the possibility it could minimise the need to handle dingoes and enable non-invasive ways of monitoring and identifying those on Fraser Island.

One of the ideas suggested by Prof Lovell is to adapt existing phone apps so that pictures could be taken and put up on social media.