Not seeing the forest for the trees? We’ve all been there.
Traditionally air rescue services have had it tough finding people who have lost their way in forests, as vegetation can obscure thermal imaging cameras.
Now a deep-learning algorithm developed by a trio of researchers may provide a solution. Images of a given area are taken by drone and processed by an AI.
These combine to form an overall picture with a far greater depth of field (like using a mega lens), meaning people on the ground are more easily spotted.
The system is 87-95% accurate, compared with a 25% accuracy rating for traditional thermal imaging.
A warming thought!