
抄録
J-013
Gesture recognition with multiple features from hands and face using depth sensor
◎Graciliano Garcia Torres Galindo Jr・Chendra Hadi SURYANTO・福井和広(筑波大)
The task of this paper is to use hand and head tracking to recognize gestures and extract informations to extend the interaction with machines. These informations will also be used together, observing the relative position between both hands and head. This will increase the set of possible gestures that can be recognized. The experiments are done with real-time recognition using a depth sensor such as Kinect and SoftKinectic DepthSense 325, apturing sequences of depth images from the device's camera. The image rocessing will be done using the OpenCV library. With this, it will be possible to experiment more intuitive human interface, being able to control any kind of application.