地球信息科学学报 ›› 2023, Vol. 25 ›› Issue (7): 1325-1335.doi: 10.12082/dqxxkx.2023.230085

• • 上一篇    下一篇

面向多源遥感影像数据的溯源模型研究

吴敏1(), 张明达1,*(), 李盼盼2, 张勇健1   

  1. 1.湖北大学资源环境学院, 武汉 430062
    2.武汉市测绘研究院, 武汉 430022
  • 收稿日期:2023-02-23 修回日期:2023-04-18 出版日期:2023-07-25 发布日期:2023-06-30
  • 通讯作者: *张明达(1987— ),男,山东济南人,副教授,硕士生导师,主要从事地理信息服务、空间数据溯源、地理模型共享与耦合等方面的研究。E-mail: mdzhang@hubu.edu.cn
  • 作者简介:吴 敏(2000— ),女,湖南醴陵人,硕士生,研究方向为地图学与地理信息科学。E-mail: wumin@stu.hubu.edu.cn
  • 基金资助:
    湖北省重点研发计划项目(2020AAA004);区域开发与环境响应湖北省重点实验室(湖北大学)开放课题(2020(B)002)

Research on Provenance Model for Multi-source Remote Sensing Images

WU Min1(), ZHANG Mingda1,*(), LI Panpan2, ZHANG Yongjian1   

  1. 1. Faculty of Resources and Environmental Science, Hubei University, Wuhan 430062, China
    2. Wuhan Geomatics Institute, Wuhan 430022, China
  • Received:2023-02-23 Revised:2023-04-18 Online:2023-07-25 Published:2023-06-30
  • Contact: *ZHANG Mingda, E-mail: mdzhang@hubu.edu.cn
  • Supported by:
    Key R&D Program of Hubei Province, China(2020AAA004);Hubei Key Laboratory of Regional Development and Environmental Response (Hubei University)(2020(B)002)

摘要:

遥感卫星技术的快速发展推动了多层次、多角度、全方位、全天候的对地观测,遥感影像数据极大丰富了起来,已广泛应用于众多领域。然而,影像数据时空分辨率不一、质量参差不齐,从原始遥感数据到数据产品的处理算法多样、业务流程复杂,导致遥感数据处理结果的质量追溯较为困难。针对此,本文对多源遥感影像数据溯源信息模型进行研究,根据遥感影像的分发、处理等特点,构建了遥感影像元数据溯源信息概念模型,图谱化表示遥感影像衍生过程中的事件、实体、关系和属性信息;为了提高遥感影像溯源信息的互操作能力,对W3C(万维网联盟) PROV溯源模型进行了继承与扩展,构建了溯源信息概念模型与W3C PROV溯源模型的映射框架,实现了溯源信息的表达;提出了遥感影像元数据模型扩展方法,以支持溯源信息的嵌入。本研究丰富了遥感影像元数据的溯源信息,支持用户对遥感数据产品追踪溯源,评估其可用性和可靠性;同时,提出了基于溯源信息的遥感数据查询优化方法,提高了数据查询效率。

关键词: 数据溯源, 遥感影像, 溯源模型, 互操作, PROV-DM模型, 遥感数据质量, 遥感影像元数据, 溯源信息查询

Abstract:

Over the past few decades, the rapid development of satellite technologies has led to significant advancements in multi-level, multi-angle, all-directional, and all-weather observation of the Earth. Remote sensing data have been greatly enriched and widely used in many scientific fields and provide important source information for geospatial analysis, environmental monitoring, disaster management, and etc. However, it is difficult to guarantee the data reliability and availability due to the inconsistency in data quality and difference in spatial and temporal resolutions. Additionally, a single remote sensing image can also be processed using different algorithms, and the process chain for generating a remote sensing product is often complex, making it difficult to determine the remote sensing product quality. To address these issues, this paper focuses on the provenance information model for multi-source remote sensing images that allows users to track the origin, derivation, and processing history of a remote sensing product. By embedding provenance information into remote sensing image metadata, users can evaluate the quality, reliability, and availability of remote sensing data products. To achieve this goal, this paper proposes a conceptual provenance model based on the distribution and processing of remote sensing images. This conceptual model includes four key elements (event, entity, relationship, and attribute) that are associated with the data derivation. Secondly, in order to improve the interoperability, this paper inherits and extends the W3C PROV-DM provenance model, and constructs a mapping framework between the remote sensing image provenance conceptual model and the W3C PROV-DM model. Furthermore, this paper enriches the metadata model and Unified Metadata Model (UMM) model with provenance information. This enhanced metadata model can provide users with a more comprehensive understanding of remote sensing data products. At the same time, this paper proposes a method for optimizing remote sensing data queries based on provenance information, which can effectively reduce the query time. This method enables users to filter out irrelevant data and retrieve desired data quickly and efficiently, leading to more efficient use of remote sensing data products. In summary, the proposed provenance information model enables users to track the provenance information of remote sensing data products, evaluate their availability and reliability, and optimize the query process. These contributions are of great significance for evaluating the accessibility and reliability of remote sensing image products.

Key words: data provenance, remote sensing images, provenance model, interoperability, PROV-DM model, remote sensing data quality, remote sensing image metadata, provenance information query