HumTouch: Localization of Touch on Semi-Conductive Surfaces by Sensing Human Body Antenna Signal

HumTouch is a touch sensing technology utilizing the environmental electromagnetic wave. The method can be realized using conductive and semi-conductive materials by simply attaching electrodes to the object’s surface. In this study, we compared three methods for localizing a touch on <inline-for...

Full description

Bibliographic Details
Main Authors: Tzu Hsuan Hsia, Shogo Okamoto, Yasuhiro Akiyama, Yoji Yamada
Format: Article
Language:English
Published: MDPI AG 2021-01-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/21/3/859
Description
Summary:HumTouch is a touch sensing technology utilizing the environmental electromagnetic wave. The method can be realized using conductive and semi-conductive materials by simply attaching electrodes to the object’s surface. In this study, we compared three methods for localizing a touch on <inline-formula><math display="inline"><semantics><mrow><mn>20</mn><mo>×</mo><mn>16</mn><mspace width="3.33333pt"></mspace><msup><mrow><mi>cm</mi></mrow><mn>2</mn></msup></mrow></semantics></math></inline-formula> and <inline-formula><math display="inline"><semantics><mrow><mn>40</mn><mo>×</mo><mn>36</mn><mspace width="3.33333pt"></mspace><msup><mrow><mi>cm</mi></mrow><mn>2</mn></msup></mrow></semantics></math></inline-formula> papers, on which four or eight electrodes were attached to record the voltages leaked from the human fingertip. The number and positions of the electrodes and the data processing of the voltages differed according to the localization methods. By constructing a kernel regression analysis model between the electrode outputs and the actual physical locations, the touched locations were estimated. Each of the three methods was tested via leave-one-out cross validation. Out of the three methods discussed, two exhibited superior performances in terms of the estimation errors. Of these two methods, one simply uses the voltages recorded by the four electrodes attached on the middle of paper edges as inputs to the regression system. The other uses differential outputs of electrode pairs as the inputs. The smallest mean location errors were 0.31 cm on <inline-formula><math display="inline"><semantics><mrow><mn>20</mn><mo>×</mo><mn>16</mn><mspace width="3.33333pt"></mspace><msup><mrow><mi>cm</mi></mrow><mn>2</mn></msup></mrow></semantics></math></inline-formula> paper and 0.27 cm on <inline-formula><math display="inline"><semantics><mrow><mn>40</mn><mo>×</mo><mn>36</mn><mspace width="3.33333pt"></mspace><msup><mrow><mi>cm</mi></mrow><mn>2</mn></msup></mrow></semantics></math></inline-formula> paper, which are smaller than the size of a fingertip.
ISSN:1424-8220