A form of machine learning, that provides real-time image analysis for identifying humans in rural areas, will soon be used to improve Police Scotland’s use of remotely piloted aircraft systems (RPAS) to find missing and vulnerable people.

The technology has been developed by a consortium of partners – CENSIS, Thales UK, University of the West of Scotland and Police Scotland. It identifies where a human being is located, rather than an individual, so no facial recognition technology is involved.

The system is trained with hundreds of hours of footage of police officers in different clothing, positions, and situations. The AI then scours an image and can locate a person within seconds at a distance of up to 150 metres. The development team explain that its ability to recognise a human will improve the more it is used.

Core AI development work is complete and trials of the new system underway, so the project team expect the technology to be deployed in searches for missing and vulnerable people in the near future.

The specifically designed algorithms developed in this project can be used on a smartphone or tablet connected to an RPAS. That compares to existing technology that tends to be immobile due to requiring computers with very large amounts of data processing power.

It will initially be employed in the search for missing and vulnerable people, but the project team believe the technology could be used in a variety of other applications, including monitoring wildlife on land and at sea.

Author: Charlotte Hathway