HCI gesture tracking using wearable passive tags

Abstract

In this thesis. a wearable system is developed to track hand gestures with passive RFID sensor tags. This system was composed of an ultra-high frequency reader and small, passive, finger-worn tags powered by scavenged RFID energy equipped with a variety of sensors that could be used to detect gestures. The primary physical goals of the system were to be comfortable and wearable without interfering with other everyday activities. The computational goals of the system were to track particular hand movements, which could be used to control a wearable computer or aid in interaction with ubiquitous and other wearable devices. As a user is walking through their environments, we aim to avoid the need for pulling out an interface with keyboard, keypad, or touch screen, and also avoid bulky hand-held interfaces, allowing the user to specify input with their fingers without taking their eyes and attention off their immediate focus. This thesis first introduces our hardware, then gives some example user interface implementations, such as a mouse controlled by hand position and a click controlled by finger proximity, entering input by touching fingers., setting options when moving the hand to a particular spot of the users apparel labeled with a passive RFID tag, and otherwise mapping control onto motion of the hand, arm, and fingers. We evaluate the effectiveness of our interface via a series of user studies. The overall system was somewhat successful, but as this is an early implementation, it was still very much limited by transmit power and antenna efficiency, due to the constraints on the size of the passive tags.