Design of Underwater Acoustic Touchscreen Based on Deep Convolutional Neural Network

With the improvement of living standards, the way of human-computer interaction has gradually changed from the original mouse and keyboard to touchscreen. Unfortunately, most traditional touchscreens lose their accuracy when working in water due to the influence of the fluid-solid interface. In this paper, based on the theory of leaky-Lamb wave, an acoustic touchscreen used underwater is proposed. The proposed acoustic touchscreen device is composed of three piezoelectric transducers and a sheet of quartz glass. Pressing position and pressing pressure on this underwater touchscreen can be successfully identified by the trained convolutional neural network. According to the experiments, the recognition accuracy of the pressing position of the testing set is 96.90 % with the spatial resolution of 1 cm x 1 cm. When it comes to recognizing the pressing pressure, the recognition accuracy of the testing set is 91.84 %.(c) 2023 Elsevier Ltd. All rights reserved.