Transactions of the Society of Instrument and Control Engineers
Online ISSN : 1883-8189
Print ISSN : 0453-4654
Paper
The Development of a Noncontact Letter Input Interface “Fingual” Using Magnetic Dataset
Taishi FUKUSHIMAFumio MIYAZAKIAtsushi NISHIKAWA
Author information
JOURNALS FREE ACCESS

2012 Volume 48 Issue 3 Pages 159-166

Details
Abstract

We have newly developed a noncontact letter input interface called “Fingual”. Fingual uses a glove mounted with inexpensive and small magnetic sensors. Using the glove, users can input letters to form the finger alphabets, a kind of sign language. The proposed method uses some dataset which consists of magnetic field and the corresponding letter information. In this paper, we show two recognition methods using the dataset. First method uses Euclidean norm, and second one additionally uses Gaussian function as a weighting function. Then we conducted verification experiments for the recognition rate of each method in two situations. One of the situations is that subjects used their own dataset; the other is that they used another person's dataset. As a result, the proposed method could recognize letters with a high rate in both situations, even though it is better to use their own dataset than to use another person's dataset. Though Fingual needs to collect magnetic dataset for each letter in advance, its feature is the ability to recognize letters without the complicated calculations such as inverse problems. This paper shows results of the recognition experiments, and shows the utility of the proposed system “Fingual”.

Information related to the author
© 2012 The Society of Instrument and Control Engineers
Previous article Next article
feedback
Top