Identifying a hedgehog from space is soon a reality. But why? As Arstechnica reports, researchers at the University of Cambridge are trying to use satellite images and AI models to map the possible habitats of hedgehogs throughout the United Kingdom. For this, they identify their favorite hiding places … Rales.
An approach that responds to a crucial participation in conservation since the population of European hedgehogs has decreased from 30 to 50% in the last ten years. In October 2024, hedgehogs of Western Europe went from “minor concern” to “almost threatened” of extinction in the red list of the International Union for the Conservation of Nature (IUCN).
But the traditional processes to detect them are long and expensive. In fact, hedgas censuses require intense night work, specialized teams or even the help of scientists. Therefore, English researchers develop a more appropriate device to detect and preserve large -scale Hedgehs.
A model still in the test phase
Satellite images can cover a large continuous area. For its part, a simple AI model is able to reliably identify the habitat characteristics of these mammals, such as brambles, based on satellite images and inaturalist data, a citizen science platform where people share their fauna observations. The tool then assigns trust scores to the detected areas, depending on the probability of finding brambles.
In general, the model is more efficient to detect large non -covered braces, visible from above. The smallest, hidden under the trees, for example, have lower trust scores.
At the moment, the model is still in the trial phase. If the team’s research ends, the Bramble detector could be very useful in the field. In fact, unlike learning models such as Chatgpt, more resources, the system can work on mobile and nomadic devices.
Similar approaches based on AI, which combine satellite remotetection and scientific observations, could help map other species, track animals that devastate crops or monitor the evolution of several ecosystems.
Source: BFM TV
