Satellite Remote Sensing Information Knowledge Graph Construction And Interpretation Method Based On Target Characteristic

The project is based on the construction of satellite remote sensing information knowledge atlas and intelligent auxiliary interpretation technology of target characteristics.For remote sensing satellite image data,two plug-ins,namely,the construction of image information interpretation body library and the management of associated auxiliary interpretation knowledge atlas,have been established,including six modules,namely,the construction of ontology database,the management of ontology inventory,the map database,the identification of entity attributes,the extraction of association relations,and the knowledge fusion,The construction of satellite remote sensing information knowledge map based on target characteristics and the function construction of intelligent auxiliary interpretation technology are realized.The target database covers not less than 3 types of association analysis,provides not less than 100000 pieces of prior strategic information,and the target knowledge map covers not less than 16 types and150 types of targets.This project is based on the strategic data of retrieving the original remote sensing image from the ground system,using the knowledge map technology to associate and organize the management of the target labeling information,and building the knowledge map based on the labeling data,prior knowledge,and unstructured text data centered on aircraft,vehicles,ships,and other targets,and finally forming an intelligent visual display.Through the application and development of knowledge map construction and intelligent aided interpretation technology,the efficiency of traditional remote sensing sample labeling has been improved.Compared with the traditional labeling methods relying on expert experience,from the perspective of target interpretation and image recognition,the construction of satellite remote sensing information knowledge map based on target characteristics and intelligent auxiliary interpretation technology can improve the accuracy of remote sensing image data labeling,reduce the technical difficulty of identifying types and models in the process of target labeling,and further promote the development and progress of aerospace industry.