Toward the reduction of incorrect drawn ink retrieval
Atsushi Kitani, Taketo Kimura, Takako Nakatani · Human-centric Computing and Information Sciences · 2017
Abstract As tablet devices become popular, various handwriting applications are used. Some of applications incorporate a specific function, which is generally called palm rejection. Palm rejection enables application users to put the palm of a writing hand onto a touch display. It classifies intended touches and unintended touches so that it prevents accidental inking, which has been known to occur under the writing hand. Though some of palm rejections can remove accidental inking afterward, this function occasionally does not execute correctly as it may remove rather correct ink strokes as well. We call this interaction Incorrect Drawn Ink Retrieval (IDIR). In this paper, we propose a software algorithm that is a combination of two palm rejection logics that reduces IDIR with precision and without latency. That algorithm does not depend on specific hardware, such as an active stylus pen. Our data provides 98.98% correctness and the algorithm takes less than 10 ms for the distinction. We confirm that our experimental application reduced the occurrences of IDIR throughout an experiment.