Orders with manual signs
Text created by automatic translator Elia and has not been subsequently revised by translators. Elia Elhuyar
In
Germany, the University of Duisburg and the Imech Institute have worked in the handling of heavy machines. A gesture-recognition system has been installed on the 7-metre-long tunnel drilling machine. The neural network and the statistical procedure are the essence of this “intelligent” system.
The machine has had a learning with about twenty people and already knows how to distinguish 24 different gestures, regardless of the environment. The drilling arm and the hydraulic control work automatically, collecting and interpreting the driver's gestures through the camera. The probability of correctly interpreting the 24 gestures studied is 93% and 10 well interpreted gestures of 100%.
Zu idazle
Zientzia aldizkaria
