Australia is legendary for just a few issues: sunshine, wonderful seashores and an enormous array of lethal animals that can chunk/poison/sting you at a second’s discover.
With that in thoughts, a gaggle of researchers have created a brand new shark-detecting drone able to find the apex predators underwater, faster than the human eye and with the next stage of accuracy.
The know-how, often called SharkSpotter, makes use of an algorithm to detect sharks in a stay video feed recorded in actual time by a drone (often called the Little Ripper Livesaver) flying above the water. Utilizing a world-first algorithm, developed utilizing synthetic intelligence and deep neural networks, SharkSpotter is ready to distinguish sharks from dolphins, rays and different marine animals, and even surfers. Because of an onboard megaphone, the drone also can warn swimmers about what’s lurking within the water earlier than they’ve even seen the risk.
Australia noticed a complete of 26 shark assaults in 2016, together with two fatalities, in line with statistics from Taronga Conservation Society Australia. However whereas that quantity is arguably low (significantly in comparison with the variety of swimmers that hit Australia’s seashores each summer time) the spectre of shark assaults looms massive over the Australian consciousness.
However the brand new know-how has the potential to extra precisely detect the animals earlier than a possible shark-human encounter even turns into a threat.
The drones are the brain-child of The Little Ripper Group, which labored in partnership with researchers from the College of Expertise Sydney’s College of Software program for greater than a 12 months to develop the shark-spotting know-how.
UTS Professor Michael Blumenstein stated the know-how may have an actual constructive affect for the general public through the use of know-how to make seashores safer.
“The automated system for detection and identification of sharks particularly, and marine life/objects extra usually, was developed utilizing innovative deep neural networks and pictures processing strategies,” stated UTS Professor Michael Blumenstein.
“The system effectively distinguishes and identifies sharks from different targets by processing video feeds which are dynamic in addition to photos, the place objects are static.”
The drones will start common patrols alongside seashores in New South Wales and Queensland on Australia’s east coast from September, for the beginning of the Surf Life Saving patrol season.
Tech Tradition: From movie and tv to social media and video games, here is your home for the lighter facet of tech.
Batteries Not Included: The CNET crew shares experiences that remind us why tech stuff is cool.