Improved Morse Code Recognition and Real-Time Translation System Based on a Low-Cost, Tailorable Flexible Capacitive Sensor
Feilu Wang, Hao Wang, Lang Wu, Tongjie Liu, Renting Hu, Yang Song
Abstract
The increasing demand for flexible wearable electronic devices has prompted the rapid development of pressure sensors capable of monitoring a range of human movements and physiological signals. However, an increasing number of studies are requiring pressure sensors with high performance while also seeking low-cost, large-scale, or even disposable manufacturing methods. In this study, we propose a flexible capacitive sensor based on low-cost, tailorable materials. The sensor employs a polyurethane sponge coated with inert metals, namely copper and nickel (Cu@Ni/PUS), as electrodes and a polyimide (PI) film as the dielectric layer. The prepared pressure sensors exhibit high sensitivity (0–17.5 kPa, 49.14% kPa –1 ), a rapid response time (80 ms), low hysteresis (6.49%), and high stability. Furthermore, the Cu@Ni/PUS sensor was integrated into an insulated glove, and an innovative, improved Morse code encoding scheme was developed. By combining the CNN-TCN dual-channel neural network model, we achieved the high-precision classification of the 26 alphabet letters (99.20%), providing a method for low-cost, high-efficiency Morse code transmission using flexible tactile sensors. Based on these findings, we also developed an improved Morse code real-time translation system that completes real-time encryption and decryption of improved Morse code gestures. The experimental results demonstrate that this research has broad application prospects in low-cost, accessible human–computer interaction and wearable devices for individuals with disabilities.