Cross Modal Retrieval Method for Text Labels of Substation Drawings Based on Deep Hashing Algorithm
Traditional information retrieval methods mainly rely on single mode data processing,which is difficult to effectively deal with the complexity and heterogeneity of such multimodal data.How to efficiently achieve cross modal retrieval between substation drawings and text labels?Therefore,this paper proposes a cross modal retrieval method for text labels of substation drawings based on deep hash algorithm.By extracting text elements from substation drawings,their specific location information in the substation drawings is reflected,laying a solid foundation for cross modal retrieval of text labels in subsequent substation drawings.Graph convolutional network technology is used to extract row depth features from the retrieved labels,generate tag embedding vectors rich in semantic relevance,and perform retrieval prediction with the corresponding text of substation drawings.
Text TagsPower SystemSubstation DrawingsCross Modal RetrievalDeep Hashing Algorithm
万方数据
维普
