{"id":564959,"date":"2026-03-26T05:23:17","date_gmt":"2026-03-26T05:23:17","guid":{"rendered":"https:\/\/www.newsbeep.com\/au\/564959\/"},"modified":"2026-03-26T05:23:17","modified_gmt":"2026-03-26T05:23:17","slug":"ai-crocodile-detection-trials-begin-in-north-queensland","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/au\/564959\/","title":{"rendered":"AI crocodile detection trials begin in north Queensland"},"content":{"rendered":"

Researchers are hoping a trial of artificial intelligence (AI) that automatically detects crocodiles could lead to an early alert system in north Queensland waters.<\/p>\n

The prototype \u2014 developed by the Department of Environment, Tourism, Science and Innovation (DETSI) and James Cook University (JCU) \u2014 uses AI to scan waterway footage live streamed from a camera mounted on a movable trailer.<\/p>\n

Over the next year, there will be three week-long, proof-of-concept field trials to test its accuracy.<\/p>\n

\"Man<\/p>\n

Daniel Guymer at the Queensland Department of Science, Environment, and Innovation operations Centre in Cairns. (ABC News: Conor Byrne)<\/p>\n

The trials will take place in northern Queensland, but DETSI senior conservation officer Daniel Guymer said the department was still determining suitable test locations.<\/p>\n

He said crocodile density was a factor.<\/p>\n

“[Somewhere] we’re going to get a solid number of movements through the array to be able to glean the effectiveness of the technology,” he said.<\/p>\n

DETSI has provided $359,000 towards the project since January 2025, as well as thousands of videos to train the AI.<\/p>\n

JCU researcher and project lead Tao “Kevin” Huang said the trials would shape further improvements.<\/p>\n

Dr Huang said the system had already achieved “reasonable” results, and he was hoping for at least 70 per cent accuracy in the trials.<\/p>\n

“There are some challenges in this research because of the light reflecting from the water surface, the changing conditions, and also the floating logs and debris that look similar to crocodiles,” he said.<\/p>\n

\"Corporate<\/p>\n

Tao “Kevin” Huang says success would mean big benefits for safety. (Supplied: James Cook University)<\/p>\n

AI vs human eye<\/p>\n

Dr Huang said AI could outperform humans, but there was no guarantee of its effectiveness.<\/p>\n

“This system is designed to support early detection, trying to provide an additional layer of monitoring, but \u2026 people should always follow the existing safety advice.”<\/p>\n

The project is one of several innovations DETSI is exploring to enhance crocodile management.<\/p>\n

“It is hoped that one day, the cameras could be paired with another underwater detection method, such as potentially sonar,” Mr Guymer said.<\/p>\n

“It’s never going to replace people, and the onus for an individual’s safety is always going to rest with themselves.”<\/p>\n

\"A<\/p>\n

The AI crocodile detection system is mounted on a mobile camera trailer. (Supplied: JCU)<\/p>\n

Small needle in big haystack<\/p>\n

Barry Brook from the University of Tasmania’s Biological Sciences faculty is not involved in this research but has developed his own AI system to identify other species.<\/p>\n

He has also researched crocodiles with Charles Darwin University in Kakadu.<\/p>\n

Professor Brook said the DETSI and JCU project was typical of object-detection technology, which was becoming widely used in ecology.<\/p>\n

\"A<\/p>\n

Barry Brook uses AI for species detection and also has crocodile research experience in Kakadu National Park. (Supplied: J. Buettel)<\/p>\n

He said it was important for the project to collect plenty of “training data”, so the object detection could cope with flooding, debris and different weather conditions at a body of water.<\/p>\n

“The more [data] variety, the more of the total distribution of possible things that the camera could see are covered in that training data, the more reliable it’s going to be,” he said<\/p>\n

“If you don’t [get that variety of data], then there’s a risk that the model will lock onto things such as the colour or turbidity of the water.<\/p>\n

“But provided your training data isn’t biased like that, then the system’s very robust to those sorts of changes.”<\/p>\n

\"Corporate<\/p>\n

Ferdous Sohel says the training data will be important to the model’s success. (Supplied: Murdoch University)<\/p>\n

AI ‘not foolproof’<\/p>\n

Ferdous Sohel, a professor of AI at Murdoch University, has been involved in many similar projects using AI to detect marine life, farm animals, pests, and even ants and microorganisms.<\/p>\n

He said the project was worthwhile but not foolproof.<\/p>\n

“If we say that the crocs are coming to these waters, say 100 times, and we are able to detect them 90 per cent of the time, that’s still a lot of safety,” Professor Sohel said.<\/p>\n

He said it was important the model triggered human intervention for confirmation should a crocodile be detected, and stressed that the training data must be thorough.<\/p>\n

“If once or twice you miss [the object], the whole algorithm is a failure,” he said.<\/p>\n

He said potential issues of sun glare, wind ripples, and brown water could be solved.<\/p>\n

“Pre-processing techniques are able to clear the image so that any object of interest underneath or in the murky water is visible,” he said.<\/p>\n

Dr Huang said trial success could be hugely beneficial for safety.<\/p>\n

“It could lead to the technology being used at public locations such as boat ramps or in marinas,” he said.<\/p>\n","protected":false},"excerpt":{"rendered":"Researchers are hoping a trial of artificial intelligence (AI) that automatically detects crocodiles could lead to an early…\n","protected":false},"author":2,"featured_media":564960,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[26],"tags":[254,64,63,11630,278205,220552,5249,14798,31801,90242,134070,12804,34100,22561,90243,46,237326,128,13749,338],"class_list":["post-564959","post","type-post","status-publish","format-standard","has-post-thumbnail","category-wildlife","tag-artificial-intelligence","tag-au","tag-australia","tag-crocodile","tag-crocodile-detection","tag-crocodylus-porosus","tag-darwin","tag-detsi","tag-far-north-queensland","tag-fnq","tag-gimuy","tag-hobart","tag-james-cook-university","tag-kimberley","tag-marlin-coast","tag-perth","tag-regional-queensland","tag-science","tag-tropical-queensland","tag-wildlife"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/564959","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=564959"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/564959\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/564960"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=564959"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=564959"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=564959"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}