Multi-Touch Tabletop System Using Infrared Image Recognition for User Position Identification

Suto, Shota; Watanabe, Toshiya; Shibusawa, Susumu; Kamada, Masaru

doi:10.3390/s18051559

Cited by 6 publications

(7 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From Equation (7), the relationship between the boundary impedance λ  and conductivity distribution σ  can be written as www.advmattechnol.de…”

Section: Machine Learning Methodsmentioning

confidence: 99%

“…Touch panel has been a must‐have component for interactive displays. Broadly utilized techniques in touch panels are resistive, [ 1,2 ] capacitive, [ 3–5 ] optical, [ 6,7 ] acoustic‐based architectures. [ 8,9 ] Among them, capacitive touch panel dominates market [ 10 ] due to its merits such as high transparency, multitouch support, low cost and detection by contact and noncontact means.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Novel Touch Interface with Ultrahigh Optical Transmittance Based on Electrical Impedance Tomography for Interactive Displays

Gao

Sun

2021

Adv Materials Technologies

View full text Add to dashboard Cite

silver nanowire (AgNW)-poly(3,4-ethylene dioxythiophene):poly (styrene sulfonate) (PEDOT:PSS) composite transparent flexible electrodes with 96% transmittance at 550 nm. [19] Nevertheless, current techniques for high transparency electrodes require cleanroom fabrication conditions, limiting their successful use in commercial products. Furthermore, some of them still blocks 10%-3% light intensity during the visible spectrum, [11,24] indicating the need for extra power consumption to maintain the display performance compared to their counterparts without touch panels installed, which shortens battery's lifetime for portable smart devices, giving rise to unpleasant user experience.In this work, to achieve ultrahigh transparency while maintaining low material cost and a simple fabrication process, an electrical impedance tomography (EIT)-based technique is presented.EIT is a low-cost and easy-to-implement detecting technology that allows real-time imaging. [25][26][27] In previous works, EIT techniques have been applied in diverse applications. In 2013, Cao and Xu applied Calderon's method to a chest-like sensing region and utilized a direct algorithm for image reconstruction. [28] They successfully reconstructed a male normal breath chest image with a 16-electrode system. In the same year, Tawil et al. proposed an impedance measurement-based artificial skin for robotics. The artificial skin was constructed using four different layers. The top layer was made of soft suede. The second layer was a highly conductive fabric. The third was a flexible carbon-loaded layer, with 19 chromium-plated brass eyelets electrodes located at its edge. The bottom layer was made of a 5 mm thick open-cell polyurethane foam. In their work, the LogitBoost algorithm was used to classify the modality of eight different touch events (tap, pat, push, stroke, scratch, slap, pull, and squeeze), the accuracy can reach 71%. [29] In 2015, Zhang et al. proposed a wearable armband sensor (Tomo [30] ), which used an eight evenly spaced copper electrodes based structure and the support vector machine (SVM) algorithm to recognize eleven hand gestures, achieving an accuracy of 96.6%. In 2018, Russo et al. [31] proposed a 16-electrode-based three-layered tactile sensor. The top is a piezoresistive fabric layer with high resistance. The second layer is a honeycomb mesh and the bottom layer is based on highly conductive fabric material. By employing a quadratic discriminant analysis (QDA), the position recognition error of this sensor ranges from 1 to 10 mm, depending on the contact locations.In traditional touch interfaces, electrodes will block the light intensity of the display, so extra power is required to maintain the display performance. In this article, an eight-electrode electrical impedance tomography system for touch location detection is proposed to improve the transmittance. The electrodes are settled at the edge of the touch panel and a machine learning algorithm is used to perform the regression process. This system can achieve almost 1 mm m...

show abstract

“…From Equation (7), the relationship between the boundary impedance λ  and conductivity distribution σ  can be written as www.advmattechnol.de…”

Section: Machine Learning Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Novel Touch Interface with Ultrahigh Optical Transmittance Based on Electrical Impedance Tomography for Interactive Displays

Gao

Sun

2021

Adv Materials Technologies

View full text Add to dashboard Cite

show abstract

“…This method has been developed for years and is considerably mature and reliable. However, the number of studies focusing on transforming daily objects into touch-sensitive objects has been increasing [ 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 ]. Acoustic waves are used for localizing touch on rigid planes [ 4 , 5 ].…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, these materials are often used as pressure sensors and flexion sensors. Cameras are also used for touch localization [ 11 , 12 , 13 , 14 , 15 ]. This optical method does not depend on the surface material but is limited owing to the issue of occlusion.…”

Section: Introductionmentioning

confidence: 99%

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

Hsia

Okamoto

Akiyama

et al. 2021

Sensors

View full text Add to dashboard Cite

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 20×16cm2 and 40×36cm2 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 20×16cm2 paper and 0.27 cm on 40×36cm2 paper, which are smaller than the size of a fingertip.

show abstract

“…In shape recognition tasks, the pressure information collected by the sensor array is treated as an image for subsequent processing [ 4 , 5 , 6 ]. Thus, many artificial design approaches that achieve outperformance in the visual domain have been transferred to tactile shape perception.…”

Section: Introductionmentioning

confidence: 99%

A Novel Bilinear Feature and Multi-Layer Fused Convolutional Neural Network for Tactile Shape Recognition

Jie

Cai

et al. 2020

Sensors

View full text Add to dashboard Cite

Convolutional neural networks (CNNs) can automatically learn features from pressure information, and some studies have applied CNNs for tactile shape recognition. However, the limited density of the sensor and its flexibility requirement lead the obtained tactile images to have a low-resolution and blurred. To address this issue, we propose a bilinear feature and multi-layer fused convolutional neural network (BMF-CNN). The bilinear calculation of the feature improves the feature extraction capability of the network. Meanwhile, the multi-layer fusion strategy exploits the complementarity of different layers to enhance the feature utilization efficiency. To validate the proposed method, a 26 class letter-shape tactile image dataset with complex edges was constructed. The BMF-CNN model achieved a 98.64% average accuracy of tactile shape. The results show that BMF-CNN can deal with tactile shapes more effectively than traditional CNN and artificial feature methods.

show abstract

Multi-Touch Tabletop System Using Infrared Image Recognition for User Position Identification

Cited by 6 publications

References 24 publications

A Novel Touch Interface with Ultrahigh Optical Transmittance Based on Electrical Impedance Tomography for Interactive Displays

A Novel Touch Interface with Ultrahigh Optical Transmittance Based on Electrical Impedance Tomography for Interactive Displays

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

A Novel Bilinear Feature and Multi-Layer Fused Convolutional Neural Network for Tactile Shape Recognition

Contact Info

Product

Resources

About