In a collaborative partnership with Fujitsu Australia and Indigenous
Precision Services, we have designed and created a system involving
'ecology-AI'
and the use of long-range drones for surveying kangaroo species, as well
as other native and introduced fauna.
WildAI was initiated on the need to create more efficient and reliable methods to survey kangaroos (macropods). The focus has been around developing tech-based methods to detect and identify kangaroos at the species level from vast amounts of spatial imagery, collected from long-range drones with high-resolution cameras.
This work has been expanded to include many native (eg. red kangaroo, emu, wallaroo, swamp wallaby) livestock (eg. cow, sheep, horse, alpaca) and feral species (eg. red fox, fallow deer, hare).
The intent is to provide efficient real-world survey applications, with the aim of providing detection and identification of multiple species, at the species-level. We are also developing post-AI output workflows, to provide abundance/density/population estimates over large areas.
