Intelligent purification of natural resource element change polygons driven by remote sensing spatiotemporal knowledge graphs

Recently, Journal of Geo-Information Science published the latest research results of Professor Li Yansheng and his team from the School of Remote Sensing and Information Engineering, Wuhan University. The research team proposed a remote sensing spatiotemporal knowledge graph-driven natural resource element change polygon purification algorithm. The algorithm addresses the high false alarm rate and the heavy reliance on manual intervention in traditional deep learning-based change detection models. The team designed a novel remote sensing spatiotemporal knowledge graph ontology model and developed an efficient spatial analysis tool for graph databases. By integrating multi-source data knowledge extraction technologies, the proposed intelligent change polygon purification method significantly reduces false alarm rates while maintaining a high recall rate, thereby improving the efficiency of natural resource element change monitoring.

The method was validated through a natural resource element change polygon purification task in Guangdong Province from March to June 2024, with results showing a true-preserved rate of 95.37% and a false-removed rate of 21.82%. This indicates that the use of remote sensing spatiotemporal knowledge graphs for change polygon purification can effectively eliminate false alarm polygons while preserving real change information, providing a more accurate and efficient solution for natural resource monitoring.

Traditional remote sensing change monitoring methods often face issues such as high false alarm rates and heavy manual intervention. This study successfully alleviates these challenges by proposing the remote sensing spatiotemporal knowledge graph-driven intelligent purification method for natural resource element change polygons. By combining intelligent reasoning through spatiotemporal knowledge graphs with multi-source data integration, the research innovatively improves the automation level of change polygon purification, significantly reducing the need for manual intervention and providing a more efficient and precise solution for natural resource monitoring. This innovative method not only offers a new technical path for remote sensing change monitoring but also strongly supports the intelligent and automated monitoring of natural resources, with broad application prospects.

For more details, please refer to the original article:

Intelligent purification of natural resource element change polygons driven by remote sensing spatiotemporal knowledge graphs.

https://www.sciengine.com/JGIS/doi/10.12082/dqxxkx.2025.240571（If you want to see the English version of the full text, please click on the “iFLYTEK Translation” in the article page.）