Satellite‐air‐ground integrated multi‐source earth observation and machine learning processing brain for tailings reservoir monitoring and rapid emergency response
Yuting Wan,Yanfei Zhong,Ailong Ma,Xin Hu,Lifei Wei
DOI: https://doi.org/10.1002/ldr.4580
2022-12-19
Land Degradation and Development
Abstract:Tailings reservoirs are an inevitable part of the production process of metal mines, which is the typical post‐mining and post‐industrial lands with waste soil and waste water. Tailings reservoirs are often located in mountainous areas, earth observation and remote sensing is a necessary means. However, the existing observation are based on satellite or air or ground platform, satellite‐air‐ground observation is isolated, the coordination ability of processing methods is insufficient. Therefore, the response to sudden major environmental and disaster events is slow and lagging. In this study, inspired by the human brain, a Satellite‐Air‐Ground integrated multi‐source earth observation and machine learning processing brain system was built up for Tailings Reservoir stereoscopic monitoring and rapid emergency response (SAGTR). Firstly, the brain‐inspired human visual multimodal data acquisition by satellite‐air‐ground integration platform has been established to break through the traditional single‐source and single‐mode data, which involves satellite remote sensing images, UAV hyperspectral images, and ground parameters. Moreover, it is extremely unique in the current tailings reservoir monitoring system. In addition, for processing methods, break through the traditional model driven methods and build the brain‐inspired visual cortex based data‐driven automatic machine learning interpretation methods, such as deep learning classification and fast change detection. Finally, from large‐scale monitoring of natural resources to environmental emergency response in key locations, and from qualitative analysis to quantitative decision‐making. In general, the proposed SAGTR framework has outstanding significance in improving the comprehensive monitoring and emergency response capacity of tailings reservoir. This article is protected by copyright. All rights reserved.
soil science,environmental sciences