Research on Text Conversion Method for Multi-modal Data
The original way to solve machine fault diagnosis by human is inefficient and challenging,and it can no longer satisfy the power plant's demand for production.OCR(Optical Character Recognition)is an effective technique for detecting and diagnosing power fault.This paper summarizes the recognition process and the application of OCR in the field of electric power,in order to promote the understanding of this technology and explore the unsolved problems.This paper covers the research and application of interactive text extraction method for multi-modal data,aiming at the problems of power generation equipment fault information extraction,analysis,processing speed and low efficiency existing in enter-prises in the modern industrial era.According to the research,applying OCR can significantly increase the effectiveness of mending damaged equipment and raise each company's output level.Furthermore,this paper constructs a knowledge graph for power equipment fault and realizes the intelligent management.
OCR identificationpower generation equipment maintenancemulti-modal data processing
NETL
NSTL
万方数据
