测绘学报 ›› 2024, Vol. 53 ›› Issue (6): 1165-1179.doi: 10.11947/j.AGCS.2024.20230469

• 智能化测绘 • 上一篇    下一篇

高分辨率遥感影像样本库动态构建与智能解译应用

顾海燕1,2(), 杨懿1,2, 李海涛1,2(), 孙立坚1,2, 丁少鹏1,2, 刘世琦1,2   

  1. 1.中国测绘科学研究院,北京 100036
    2.测绘科学与地球空间信息技术自然资源部重点实验室,北京 100036
  • 收稿日期:2023-10-11 发布日期:2024-07-22
  • 通讯作者: 李海涛 E-mail:guhy@casm.ac.cn;lhtao@casm.ac.cn
  • 作者简介:顾海燕(1982—),女,博士,研究员,研究方向为遥感影像智能解译与高性能计算。 E-mail:guhy@casm.ac.cn
  • 基金资助:
    国家重点研发计划(2023YFB3907600)

Dynamic construction of high-resolution remote sensing image sample datasets and intelligent interpretation applications

Haiyan GU1,2(), Yi YANG1,2, Haitao LI1,2(), Lijian SUN1,2, Shaopeng DING1,2, Shiqi LIU1,2   

  1. 1.Chinese Academy of Surveying and Mapping, Beijing 100036, China
    2.Key Laboratory of Geospatial Technology for the Surveying and Mapping Sciences of the Ministry of Natural Resources, Beijing 100036, China
  • Received:2023-10-11 Published:2024-07-22
  • Contact: Haitao LI E-mail:guhy@casm.ac.cn;lhtao@casm.ac.cn
  • About author:GU Haiyan (1982—), female, PhD, researcher, majors in intelligent interpretation and high-performance computing of remote sensing images. E-mail: guhy@casm.ac.cn
  • Supported by:
    The National Key Research and Development Program of China(2023YFB3907600)

摘要:

在人工智能时代,遥感影像解译朝着自动化智能化方向发展,高质量的样本数据集是其核心。我国积累了海量优质的时空地理信息基础数据及衍生产品,是深度学习驱动的遥感影像智能解译样本的重要来源。盘活现有数据资源,可推动人工智能与遥感解译的应用深度与广度。本文基于现有数据资源,针对样本数据集区域受限、时效性不强、类型单一等问题,研究了面向深度学习的高分遥感影像智能解译样本库动态构建技术。首先,分析了要素提取、地表覆盖分类、变化检测方面的公开样本数据集的特点,提出业务驱动的样本应需生成-动态构建-智能应用思路;其次,研究了基于历史解译成果的样本自动生成、SAM大模型提示学习引导的样本清洗精化方法及实现过程;再次,设计了具有区域性、时序性、尺度性、多传感器、多类型的样本库,以及顾及空间-时间-地类关系的动态样本数据库架构,研究了样本数据集“量化-检索-组合”动态重构过程,实现时空样本的动态管理与多维检索;最后,开展了地表覆盖分类、要素提取、变化检测等智能解译应用,验证了本文研究思路及方法的可行性,以期推动基于已有基础数据的样本数据集的有效利用,以及样本构建-管理-应用及数据-模型-业务的互联互通,为高分遥感影像智能解译样本库构建与应用提供参考思路。

关键词: 高分辨率遥感影像, 样本库, 样本精化, 动态构建, 智能解译, 深度学习, 地表覆盖分类, 变化检测

Abstract:

In the era of artificial intelligence, the interpretation of remote sensing images is moving towards automation and intelligence. High-quality sample datasets are crucial for this development. A massive amount of high-quality spatio-temporal geospatial information and derived products has been accumulated in China, which serve as an important source for deep learning-driven intelligent interpretation of remote sensing images. Leveraging existing data resources can promote the depth and breadth of artificial intelligence and remote sensing interpretation applications. Aiming at the limitations of existing sample datasets, such as regional restrictions, limited timeliness and sample types, this paper proposes a dynamic construction technique for a high-resolution remote sensing image intelligent interpretation sample datasets based on existing data resources. Firstly, a business-driven framework for sample datasets demand generation, dynamic construction, and intelligent application is proposed based on the characteristics analysis of publicly available sample datasets for feature extraction, land cover classification, and change detection. Secondly, this research investigates methods and implementation processes for sample generation based on historical interpretation results, as well as sample refinement through SAM (segment anything model) large model prompt learning-guided cleaning. Furthermore, a sample dataset with regional, temporal, scale, multi-sensor, and multi-type features is designed, along with spatial-temporal-land cover relationships. The study explores the dynamic reconstruction process of sample dataset quantification-retrieval-combination, achieving dynamic management of spatiotemporal samples and multidimensional retrieval. Finally, intelligent interpretation applications such as land cover classification, feature extraction, and change detection are conducted to validate the feasibility of the proposed methods. The aim is to promote the effective utilization of sample datasets based on existing foundational data, as well as the interconnection of sample construction-management-application and data-model-business, providing reference ideas for the construction and application of high-resolution remote sensing image intelligent interpretation sample datasets.

Key words: high-resolution remote sensing images, sample datasets, sample refinement, dynamic construction, intelligent interpretation, deep learning, land cover classification, change detection

中图分类号: