UAS with real time image recognitiono for a safer society.

Important results from the project

The purpose of the project is to create a drone system with real-time image recognition that will be able to fly programmed routes or fly manually over an industrial area outside the drone pilot´s view to be able to quickly perceive people, vehicles or fire. With a "Dashboard" function where collected information is presented in a clear and well-filtered way, the system can quickly create an overview of a situation and communicate this in real time to a command center for action. The goal of the project is to create a drone system with real-time image recognition.

Expected long term effects

The project has succeeded in developing an image recognition algorithm that can detect the objects of interest with a relatively high probability. The way the algorithm is designed means that those who today do manual image recognition can participate and "train" the algorithm in new environments. This means that the software has the potential to be used in many more areas of use outside the security industry.

Approach and implementation

Tests with drones, data transfer and image recognition algorithm including test flight have been carried out with successful results. The image recognition component still requires a lot of computing power to be able to do this in real time. The project has proven that the concept is possible but more time for training the algorithm and simplification in the algorithm´s computational ability so laptops with less processor capaicity should be able to perform image recognition in the field in real time.