Northern Tablelands Local Land Services (LLS) invasive species officer Jonathan Lawson believes there could come a time when unmanned aerial vehicles (UAVs) – otherwise known as drones – are a regular fixture on local properties, following continuing breakthroughs in invasive and dangerous weed detection systems from the University of Sydney’s Department of Robotics.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
While UAVs are becoming increasingly prevalent in a number of industry applications, Mr Lawson said the introduction of a complex algorithm that identifies invasive weed species by analysing pixel properties in aerial photography is widening the range of agricultural applications for drone and remotely controlled aircraft.
Coming off the back of eight years’ research at the University of Sydney, and $700,000 in federal government funding to trial UAV weed detection in Glen Innes and the New England region, Mr Lawson explained how the complex algorithm has been developed to analyse aerial or satellite photography, extracting individual pixels and comparing the pixel properties to previously marked weed species thereby providing a pervasive analysis of weed location and density on a property to a specific degree of surety.
With the drones proving highly successful in trials around the New England region, Mr Lawson said he is keen to start pushing the limits of what the algorithm can achieve in terms of land analysis and management.
“It is working well for a number of dangerous species, but now it is about seeing how far we can take it,” he said.
Commending the work of the Department of Robotics at the University of Sydney, Mr Lawson said the region could benefit from a continued partnership with the research centre, adapting the technology to the agricultural industry.
Suggesting the possibility of incorporating satellite images, Mr Lawson said the technology to develop a universal imaging program, covering large areas of the state and identifying a number of dangerous or invasive weed species, could be one of a number of applications of the analysis algorithm that is currently in a constant learning state providing more certain analytical information with each application.
While Mr Lawson agreed that the technology is costly now, there is great potential for future farming practices to incorporate aerial imaging technology as a regular component of property management, with the components becoming more cost-effective as development continues.
