Research and Application of Instrument Reading Recognition Algorithm Based on Deep Learning

Abstract

As an important part of China's social production, instruments are scientifically applied in many fields such as industrial development and military development, as well as in substations with complex environment. Therefore, it is necessary to ensure the accuracy of identification readings in instrument application, so as to provide real and effective data information for instrument application. Based on the connotation of deep learning, this paper proposes a pointer and digital multi and multi type instrument reading recognition algorithm based on Mask -RCNN. Then, the experimental results of instrument identification reading algorithm are systematically analyzed. Finally, the application of instrument identification reading algorithm is discussed in order to improve the scientificity and accuracy of instrument identification reading algorithm.