Interacting with Objects in the Environment by Gaze and Hand Gestures

A head-mounted wireless gaze tracker in the form of gaze tracking glasses is used here for continuous and mobile monitoring of a subject's point of regard on the surrounding environment. We combine gaze tracking and hand gesture recognition to allow a subject to interact with objects in the environment by gazing at them, and controlling the object using hand gesture commands. The gaze tracking glasses was made from low-cost hardware consisting of a safety glasses' frame and wireless eye tracking and scene cameras. An open source gaze estimation algorithm is used for eye tracking and user's gaze estimation. A visual markers recognition library is used to identify objects in the environment through the scene camera. A hand gesture classification algorithm is used to recognize hand-based control commands. When combining all these elements the emerging system permits a subject to move freely in an environment, select the object he wants to interact with using gaze (identification) and transmit a command to it by performing a hand gesture (control). The system identifies the target for interaction by using visual markers. This innovative HCI paradigm opens up new forms of interaction with objects in smart environments.