Development of a virtual keyboard based on button tracking using magnetic induction
Herman Sun, Chin‐Shyurng Fahn · Journal of the Chinese Institute of Engineers · 2007
In this paper, we present the development of a virtual keyboard based on button tracking, which consists of fingertip sensors, keyboard panels, keyboard layouts, and a control unit. The fingertip sensor is composed of a small generator coil attached on the fingertip and two sensor coils installed on the keyboard panel. This keyboard panel functions as sensor amplifiers, generator coil drivers, and as a hand coaster while a user is typing. The keyboard layout which is only a picture describing the arrangement of keys on the keyboard is used for guiding the user during the training, and the control unit is used for coordinating the keyboard system and transmitting sensors’ data to the host computer. The theoretical formulation of the fingertip sensor and the realization of the virtual keyboard are also presented. To prevent interference among the generator coils, we adopt a time division method to scan the generator and sensor coils. To obtain high flexibility in the system, training and recognition methods for identifying the keyboard keys are also proposed. At the training stage, a simple and robust vector‐quantization‐based training algorithm is developed, so that trained models of an acceptable quality can be obtained even though 25% of outliers exist in the training data. At the recognition stage, an efficient recognition algorithm based on distance measurement is also presented. Additionally, in online typing, we propose a visual feedback technique to display the corresponding keys pressed by the fingers. So far, the experimental results of the virtual keyboard as well as both the training and recognition algorithms are very satisfactory. Results reveal that our virtual keyboard is currently usable for a man‐machine interface.