Research on Braille Music Recognition Based on Convolutional Neural Network

Abstract

In this era of advanced information technology, it is increasingly essential to acquire important information in time. But there exists a wider communicational gap between visually impaired people and visually normal people. The reason is that the current recognition technology of braille is not very mature, which makes the visually impaired people unable to integrate into the current information age. In this paper, the convolutional neural network model, popular at present, is used to establish the recognition model of braille music. Compared with the previous recognition model, it can be seen as a new attempt. At the same time, we performed some basic preprocessing operations on braille music images. The results of every layer of the recognition model are also shown in detail, which is more prominent in the feature extraction of braille characters. In addition, we developed the training algorithm and test algorithm of braille music. Finally, the experimental results show that the recognition model based on convolutional neural network has good effectiveness and strong generalization ability.