British scientists have just unveiled a new tool for identifying and monitoring endangered wildlife around the world. It currently consists of an artificial intelligence system to count elephants in Africa based on satellite photos.
This technology combines satellite imagery with artificial intelligence to recognize and count elephants moving through Africa's forests and grasslands.
The first tests involve animals living in southern Africa and use images from the WorldView-3 and WorldView-4 satellites. It should be noted that this type of satellite can produce more than 5,000 sq km of images with unparalleled precision in a matter of minutes.
Elephants were chosen for these initial tests due to the fact that they are the largest land animal in the world, and therefore in theory the easiest to spot. The aim of this method is to be even more precise than the census carried out by humans, generally by low altitude planes or helicopters, but above all to do so without having to approach the animals and therefore without disturbing them in their natural environment. This technology also has no borders and can therefore follow the animals even when they move from one country to another, which is more problematic on land.
Regular monitoring is fundamental to the efforts of saving a species, the aim being to be able to map populations and anticipate their evolution. According to those in charge of this research program, the first results match up with human detection capabilities, which therefore opens up tremendous prospects in other parts of the world, where endangered species are much less monitored.