University of Queensland researcher and cowcam co-inventor, Neal Finch, explained: “We use the unique side profile that every animal has and a software programme similar to facial recognition technology that allows us to identify animals to a species level. The camera can tell the difference between sheep and cattle and feral pests such as goats, horses, pigs, kangaroos, camels and emus.
“You could have a cattle station that has feral populations of horses, donkeys or camels. The watering points are there for the cattle, so the camera would let the cattle through, but if a goat or a pig tried to get in the gate would shut against it.”
That’s when this image-processing cat door made the rounds.
this is what it sees when Flo sticks her head in from the street. The image is captured as soon as the center of the picture becomes dark. At this point our software analyses the image to determine if Flo is carrying anything in her mouth.
Of course, unless the feral animals are good enough to come alone and never zip through underneath a cow or right after it, beforethe gate closes, this thing would be a great idea.
There’s enough image tracking and processing power around that you could just dart the offending critters when they try to come in.