自然资源要素智能解译研究进展与方向

自然资源要素解译是自然资源调查监测工作中最基础、投入工作量最大、技术难度最高的工作,存在可解译类别少、要素边界不准不全、类别属性可靠性不高等诸多挑战。本文从自然资源调查监测应用场景出发,首先从要素解译到场景要素耦合的解译范式、数据驱动到知识驱动的解译方法、人工目视到人机协同的解译手段3个方面阐述了自然资源要素自动解译的研究进展;然后探讨了6个重点研究方向及其研究内容,包括场景要素耦合解译、知识驱动语义理解、人机协同智能解译、内外一体与三维环境解译、关键参数精准计算与定量反演,以及高可信质量控制与真实性验证;最后进行了总结与展望,以期为自然资源要素智能解译研究提供思路,推动解决困扰遥感影像自动解译方法在自然资源调查监测业务应用中的瓶颈问题。

Research progress and trend of intelligent interpretation for natural resources features

Interpretation of natural resources features is the most basic, most widely used, most involved, and most technically difficult task in survey and monitoring work. It faces many challenges such as application understanding, accurate interpretation, and fine interpretation. Starting from the application scenarios of natural resource survey and monitoring, firstly, this article elaborates its research progress from three aspects:scene-features coupling interpretation mode, semantic understanding expression cognition, and human-computer collaborative interpretation methods. Then, it discusses six key research directions and basic ideas, including scene-feature coupling interpretation, knowledge-driven semantic understanding, human-machine collaborative intelligent interpretation, indoor and outdoor integrated and 3D environment interpretation, quantitative inversion of key parameters, high-confidence quality control and authenticity verification. Finally, summaries are given. This paper aims to provide some ideas for the research on intelligent interpretation of natural resources survey and monitoring features, and promote the solution of bottlenecks that plague the application of remote sensing image automatic interpretation methods in natural resource survey and monitoring applications.