<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In this paper, a static gesture classification model based on a convolutional neural network is proposed to achieve the goal of efficient interaction control and fast response time for gesture recognition, and hand feature extraction is achieved by using alternating convolution and pooling. In the convolutional neural network, image data is obtained by preprocessing operations such as adaptive thresholding and functions such as Gaussian filtering to provide data for the convolutional neural network structure to build the recognition model, which is used to quantify the edge characteristics of the image and construct the image edges with gradient amplitude to build the static gesture recognition algorithm for Kinect. In the accuracy rate for gesture interaction, the average recognition rate for slow movement is 96.5%, while when the interactive gesture is recognized at an accelerated speed, its recognition rate slips to 93.1%, and the total average recognition rate reaches 95.2%.</p>
</abstract>

In this paper, a static gesture classification model based on a convolutional neural network is proposed to achieve the goal of efficient interaction control and fast response time for gesture recognition, and hand feature extraction is achieved by using alternating convolution and pooling. In the convolutional neural network, image data is obtained by preprocessing operations such as adaptive thresholding and functions such as Gaussian filtering to provide data for the convolutional neural network structure to build the recognition model, which is used to quantify the edge characteristics of the image and construct the image edges with gradient amplitude to build the static gesture recognition algorithm for Kinect. In the accuracy rate for gesture interaction, the average recognition rate for slow movement is 96.5%, while when the interactive gesture is recognized at an accelerated speed, its recognition rate slips to 93.1%, and the total average recognition rate reaches 95.2%.

Kinect-based gesture interaction design method for smart terminal APP interface

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

{"article-title":"Kinect-based gesture interaction design method for smart terminal APP interface"}

In this paper, a static gesture classification model based on a convolutional neural network is proposed to achieve the goal of efficient interaction control...

Kinect-based gesture interaction design method for smart terminal APP interface

Pubblicato online: 30 ott 2023

Pagine: -

Ricevuto: 05 dic 2022

Accettato: 10 mag 2023

DOI: https://doi.org/10.2478/amns.2023.2.00873

Parole chiave
Gestural interaction, Kinect, Static gesture recognition, Convolutional neural network, Hand feature extraction

© 2023 Hui Wang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Kinect-based gesture interaction design method for smart terminal APP interface

Pubblicato online: 30 ott 2023

Pagine: -

Ricevuto: 05 dic 2022

Accettato: 10 mag 2023

DOI: https://doi.org/10.2478/amns.2023.2.00873

Parole chiaveGestural interaction, Kinect, Static gesture recognition, Convolutional neural network, Hand feature extraction

© 2023 Hui Wang et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Parole chiave
Gestural interaction, Kinect, Static gesture recognition, Convolutional neural network, Hand feature extraction