Researchers have devised a new system that could one day allow drone aircraft taxiing on a runway to recognise and follow hand gestures.
Presently, drones can land autonomously on aircraft carrier decks, but are controlled by humans during taxiing.
With piloted aircraft, navy flight-deck marshals follow a codified set of hand gestures to direct them to, for example, cut their engines, open weapon bay doors or move to a refuelling bay.
To test whether these gestures could be recognised by a computer, Yale Song and colleagues at the Massachusetts Institute of Technology wrote an algorithm that analyses 3-second clips from a depth-sensing camera trained on a person performing flight-deck gestures, the New Scientist reported.
The system recorded body, arm, wrist, hand and finger positions and could consequently identify a flight-deck command with an accuracy of 76 per cent.
The team revealed that it is currently working towards improving recognition levels.
“I can’t see why this wouldn’t work ultimately,” said Peter van Blyenburgh, head of UVS International, a drone trade group.
“The gestures are clearly defined – an image sensor should be able to pick them up,” Blyenburgh added.
The study will appear in ACM Transactions on Interactive Intelligent Systems.
Presently, drones can land autonomously on aircraft carrier decks, but are controlled by humans during taxiing.
With piloted aircraft, navy flight-deck marshals follow a codified set of hand gestures to direct them to, for example, cut their engines, open weapon bay doors or move to a refuelling bay.
To test whether these gestures could be recognised by a computer, Yale Song and colleagues at the Massachusetts Institute of Technology wrote an algorithm that analyses 3-second clips from a depth-sensing camera trained on a person performing flight-deck gestures, the New Scientist reported.
The system recorded body, arm, wrist, hand and finger positions and could consequently identify a flight-deck command with an accuracy of 76 per cent.
The team revealed that it is currently working towards improving recognition levels.
“I can’t see why this wouldn’t work ultimately,” said Peter van Blyenburgh, head of UVS International, a drone trade group.
“The gestures are clearly defined – an image sensor should be able to pick them up,” Blyenburgh added.
The study will appear in ACM Transactions on Interactive Intelligent Systems.
No comments:
Post a Comment