For decades, the best way for conservationists to monitor threatened elephant populations has been via an aircraft survey. The problem is that elephant populations live in habitats that span international borders, which can make it difficult to obtain permission for aircraft surveys. Not to mention aircraft surveys can be expensive and time-consuming.

To help conservationists count African elephant populations, scientists in the UK have developed an algorithm that can identify elephants in satellite images.

We just present examples to the algorithm and tell it, ‘This is an elephant, this is not an elephant,’” said Dr. Olga Isupova of the University of Bath. “By doing this, we can train the machine to recognize small details that we wouldn’t be able to pick up with the naked eye.”

The scientists have tested the algorithm to look at South Africa’s Addo Elephant National Park. Through machine learning, the algorithm was able to identify elephants in a variety of backdrops, whether it be in the open savannah or in a dense cluster of trees.

According to University of Oxford scientist Dr. Isla Duporge, conservation organizations are already showing interest in using the algorithm to replace aircraft surveys, although they will have to pay for access to commercial satellites and the images they capture. That said, it should be worth it as the algorithm can survey up to 5,000 sq km of elephant habitat on a single cloud-free day.