Landsat时序变化检测综述

发布时间：2018-03-22 19:19

  本文选题：Landsat影像　切入点：时序数据　出处：《地球信息科学学报》2017年08期 　论文类型：期刊论文

【摘要】：时序变化检测已成为当前Landsat数据主流的变化检测方法。本文从检测算法对比、时序数据构建和精度评价等方面对Landsat时序变化检测进行回顾和评述,进而提出Landsat时序变化检测当前所存在的问题,及其所面临的挑战。Landsat时序变化检测算法可大致归纳为轨迹拟合法、光谱-时间轨迹法、基于模型的方法3大类,这些算法大多基于森林扰动提出;变化检测常用指标有波段型、植被指数型、线性变换型、组合型4大类,每类指标的优势不同,可综合多类指标以更全面地检测不同扰动类型。尽管Landsat时序变化检测已取得长足发展,但仍然面临诸多挑战,其中最大挑战是缺少一致性的参考数据集进行变化检测精度评价。
[Abstract]:Time series change detection has become the mainstream change detection method in Landsat data. This paper reviews and comments on Landsat timing change detection from the aspects of detection algorithm comparison, timing data construction and accuracy evaluation, etc. Furthermore, the paper puts forward the existing problems of Landsat time series change detection and its challenges. Landsat time series change detection algorithms can be roughly classified into three categories: trajectory fitting method, spectrum time track method, model-based method, etc. Most of these algorithms are based on forest disturbance. The commonly used indicators for change detection are band type, vegetation index type, linear transformation type, combined type, and the advantages of each type are different. Although Landsat time series change detection has made great progress, it still faces many challenges, among which the biggest challenge is the lack of consistent reference data set to evaluate the accuracy of change detection.
【作者单位】： 云南大学国际河流与生态安全研究院;云南省国际河流与跨境生态安全重点实验室;
【基金】：国家自然科学基金项目(41461017) 国家重点研发计划课题(2016YFA0601601) 云南省中青年学术技术带头人后备人才培育计划(2014HB005) 云南大学青年英才培育计划
【分类号】：P237
，

本文编号：1650022