This paper reports a finger-worn and portable system for Braille recognition. We contrive a brand-new tactile auditory conversion system, for the first time, by leveraging the high-density sensor array that achieves full-area scanning of Braille characters, yielding pressure distribution matrixes. A deep learning model then is developed to coding the mapping between matrixes and Braille characters, delivering real-time conversion from mechanical stimuli to electrical signals and then to audio signals. The phonic braille recognition system enables the interactive feedback between tactility and audition, which affords visually impaired individuals an enhanced degree of independence, autonomy and unhindered access to information.