Underwater acoustic 3D touch screen system based on leaked lamb waves and convolutional neural network

The invention discloses an underwater acoustic 3D touch screen system based on leaked Lamb waves and a convolutional neural network, which can effectively predict position information and pressure intensity information of a touch screen point under the condition that the surface of a screen is covered by liquid. The main structure of the device comprises a glass plate (3), a packaging water-stop sheet, a signal generator (1), a piezoelectric ceramic piece, a rear signal amplifier (6) and a jetson TX1 development plate (7). According to the device, one piezoelectric ceramic piece emits pulse waves, the other two piezoelectric ceramic pieces receive sound wave signals, original data containing contact information can be obtained, and a data set is constructed by using the original data and labels composed of positions and pressure intensities and then is used for training a convolutional neural network. And then the received original data can be directly input into the trained neural network to obtain the posit