A high-accuracy Braille recognizing sensing device bio-inspired by human touch sensation based on microstructure-based sensor and machine learning method

Braille learning is essential to communicate and work for visually impaired people (VI). However, a convenient and portable Braille learning device has not been investigated. Herein, a convenient and high-accuracy Braille learning-enabled tactile sensing system is developed to help VI learning Braille themselves. The tactile sensing system based on tailored micropatterned tactile sensor through sliding mode on Braille, which bio-inspired from human touching sensation, it achieved a high recognizing accuracy 98.96 % of 26 English letters Braille. The tactile sensor with a tiny size (3×3×2 mm) exhibited a high sensitivity of 0.11 kPa-1, which is based on two micro-dome structures at a same dimension as the dot on Braille. The tactile sensing device is fabricated by the tailored sensor, process circuit and microprogrammed control unit to recognize Braille, the sensing device can help VI learning and writing Braille. This work presents a practical theory for VI learning Braille themselves.