遥感大数据智能解译的地理学认知模型与方法

Geo-cognitive models and methods for intelligent interpretation of remotely sensed big data

随着遥感数据和计算机算力的爆炸式增长、智能分析算法瓶颈的突破,亟须提升与之相匹配的遥感大数据处理与分析能力。针对复杂场景下遥感大数据智能处理与地理学认知耦合关联和交叉融合的关键问题,本文分析了遥感大数据与地理科学各自的特点与相互关系,提出了多模态知识融合关联的深度网络构建和面向地理制图的遥感智能解译思路,建立了遥感大数据智能处理与应用体系框架;面向技术发展和行业应用,本文提出了分别建设通用高分辨率遥感智能处理系统和智能精准应用平台的总体路线,以期推动遥感智能解译技术创新和工程化应用的全面发展。

With the explosive growth of remotely sensed data and computing power,and the breakthrough of intelligent analysis algorithms,there is an urgent need to improve the capabilities to match in remotely sensed big data processing and analysis.Aiming at the crucial problems of coupling association and cross fusion of remotely sensed big data intelligent processing and geographical cognition in complex scenes,this paper analyzes the characteristics and relationships between remotely sensed big data and geographical science,puts forward the idea of building deep network of multimodal knowledge fusion and intelligent interpretation of remotely sensed data for geographical cartography,and establishes the framework of remotely sensed big data intelligent processing and application system.A general high-resolution remote sensing intelligent processing system for technology development and an intelligent application platform for industry applications are proposed,respectively,in order to boost the technological innovations and engineering applications of intelligent interpretation using remotely sensed big data.