Exact indexing of dynamic time warping
Knowledge and Information Systems
Journal of Signal Processing Systems
Handwriting analysis for diagnosis and prognosis of parkinson's disease
ISBMDA'06 Proceedings of the 7th international conference on Biological and Medical Data Analysis
Closing the loop between intentions and actions
Adjunct proceedings of the 25th annual ACM symposium on User interface software and technology
| Hi-index | 0.00 |
Handwriting dynamics which reflect fine motor skills of writers can be recorded with pen based writing systems. They are generally equipped with a diversity of sensors, such as pen tip pressure and tilt-acceleration sensors mounted inside the pen or pen tip x-y position sensors integrated on a specific graphic tablet. Such writing systems are essentially applied for biometric personal identification or handwriting recognition. In this paper, an advanced biometric pen based system for capturing and analyzing handwriting dynamics of a person is presented. Features of the device as well as evaluation of its sensor data are discussed. The system actually comprises a standard WACOM graphic tablet where its input pen is equipped additionally with a sensor to measure the grip pressure of fingers holding the pen. By combining x-y position data of the tablet and grip pressure data of the pen an improvement of performance in handwriting and person recognition is achieved. The experimental results have shown that among the single sensors, the grip sensor data gives best recognition accuracy and improves the recognition rates of handwritten PINs or persons by about 1%, when fused with x-y position data. It shows excellent accuracy in handwriting recognition and depicts detailed information about fine motor skill which is primarily because of data sampled by the finger grip pressure sensor. The enhanced input device has great promise not only for biometrics but also for biomedical applications.