基于面向对象的土地覆盖变化检测方法研究
发布时间：2018-08-26 14:08
【摘要】：土地覆盖变化检测是根据对同一物体或现象不同时间的观测来确定地表覆盖变化的处理过程,是根据不同时相的遥感图像及其它多时相GIS数据来提取、描述目标随时间产生的变化,并定量分析、确定其变化内容的理论和方法。它以GIS和RS为主要工具从多时相数据中提取人类活动和地球环境的动态信息,其结果常用于GIS空间数据库的快速更新和空间数据挖掘。 随着地理信息系统和趋于“三多”(多角度、多传感器和多平台)以及“三高”(高光谱分辨率、高空间分辨率和高时间分辨率)的遥感技术的快速发展,土地覆盖变化检测已成为空间信息科学中一个较为活跃的研究方向。面向对象影像分析技术(Object-Oriented Analysis, OOA)以对象为基本操作单元,是结合RS和GIS的一种新型的地学分析工具,广泛且成功的应用于以遥感图像分类为代表的遥感影像分析处理之中,越来越多的受到土地覆盖应用研究者的青睐。 本文在全面分析、总结国内外研究现状的基础上,介绍了变化检测的基本概念、模型和处理步骤；总结了现有的土地覆盖变化检测方法,阐述了常用方法的原理,并对它们的优缺点进行了详细的分析；研究了面向对象影像分析的相关概念和理论以及多尺度图像分割技术,并以面向对象遥感影像分析软件eCognition(易康)中的多尺度图像分割为例探讨了基于多光谱遥感图像的多尺度图像分割中参与分割的图像波段和各波段权重这两组重要参数的确定。在此基础上将面向对象遥感影像分析技术引入以Landsat5TM中分辨率多光谱遥感影像为数据源的土地覆盖变化检测中,详细对比了对象级变化检测与传统像元级变化检测的性能指标；分析探讨了三种常用变化检测方法的优劣以及面向对象土地覆盖变化检测的优势；结合变化检测的发展趋势,将LBV变换与传统的变化矢量分析(CVA)相结合,提出了一种基于面向对象的LBV-CVA变化检测法。通过实验,证明该方法能有效提高土地覆盖变化检测的精度,降低检测误差。最后对本文的研究成果进行了总结,指出了在研究中有待改进的地方,对以后的研究工作进行了展望。
[Abstract]:Land cover change detection is a processing process to determine land cover change based on the observation of the same object or phenomenon at different times. It is extracted from remote sensing images and other multitemporal GIS data. Describe the change of target over time, and determine the theory and method of change content. It uses GIS and RS as the main tools to extract the dynamic information of human activities and the earth environment from the multitemporal data. The results are often used in the rapid updating of GIS spatial database and spatial data mining. With the rapid development of geographic information systems and remote sensing technologies that tend to be "multi-angle, multi-sensor and multi-platform" and "three high" (hyperspectral resolution, high spatial resolution and high temporal resolution), Land cover change detection has become an active research direction in spatial information science. Object oriented image analysis (Object-Oriented Analysis, OOA) is a new geoscience analysis tool combined with RS and GIS. It is widely and successfully applied to remote sensing image analysis and processing represented by remote sensing image classification. More and more researchers are interested in land cover application. On the basis of comprehensive analysis and summarization of current research situation at home and abroad, this paper introduces the basic concepts, models and processing steps of change detection, summarizes the existing methods of land cover change detection, and expounds the principles of common methods. The advantages and disadvantages of them are analyzed in detail, and the related concepts and theories of object oriented image analysis and multi-scale image segmentation technology are studied. Taking the multi-scale image segmentation in object oriented remote sensing image analysis software eCognition as an example, the determination of the two important parameters of image band and the weight of each band in multi-scale image segmentation based on multispectral remote sensing image is discussed. On this basis, the object oriented remote sensing image analysis technology is introduced into the land cover change detection based on Landsat5TM resolution multispectral remote sensing image, and the performance indexes of object level change detection and traditional pixel level change detection are compared in detail. The advantages and disadvantages of three commonly used change detection methods and the advantages of object-oriented land cover change detection are analyzed and discussed, and the LBV transform is combined with the traditional change vector analysis (CVA) according to the development trend of change detection. An object-oriented LBV-CVA change detection method is proposed. The experimental results show that the method can effectively improve the accuracy of land cover change detection and reduce the detection error. Finally, this paper summarizes the research results, points out the areas to be improved in the research, and looks forward to the future research work.
【学位授予单位】：河南工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208;P237
本文编号：2205103
[Abstract]:Land cover change detection is a processing process to determine land cover change based on the observation of the same object or phenomenon at different times. It is extracted from remote sensing images and other multitemporal GIS data. Describe the change of target over time, and determine the theory and method of change content. It uses GIS and RS as the main tools to extract the dynamic information of human activities and the earth environment from the multitemporal data. The results are often used in the rapid updating of GIS spatial database and spatial data mining. With the rapid development of geographic information systems and remote sensing technologies that tend to be "multi-angle, multi-sensor and multi-platform" and "three high" (hyperspectral resolution, high spatial resolution and high temporal resolution), Land cover change detection has become an active research direction in spatial information science. Object oriented image analysis (Object-Oriented Analysis, OOA) is a new geoscience analysis tool combined with RS and GIS. It is widely and successfully applied to remote sensing image analysis and processing represented by remote sensing image classification. More and more researchers are interested in land cover application. On the basis of comprehensive analysis and summarization of current research situation at home and abroad, this paper introduces the basic concepts, models and processing steps of change detection, summarizes the existing methods of land cover change detection, and expounds the principles of common methods. The advantages and disadvantages of them are analyzed in detail, and the related concepts and theories of object oriented image analysis and multi-scale image segmentation technology are studied. Taking the multi-scale image segmentation in object oriented remote sensing image analysis software eCognition as an example, the determination of the two important parameters of image band and the weight of each band in multi-scale image segmentation based on multispectral remote sensing image is discussed. On this basis, the object oriented remote sensing image analysis technology is introduced into the land cover change detection based on Landsat5TM resolution multispectral remote sensing image, and the performance indexes of object level change detection and traditional pixel level change detection are compared in detail. The advantages and disadvantages of three commonly used change detection methods and the advantages of object-oriented land cover change detection are analyzed and discussed, and the LBV transform is combined with the traditional change vector analysis (CVA) according to the development trend of change detection. An object-oriented LBV-CVA change detection method is proposed. The experimental results show that the method can effectively improve the accuracy of land cover change detection and reduce the detection error. Finally, this paper summarizes the research results, points out the areas to be improved in the research, and looks forward to the future research work.
【学位授予单位】：河南工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208;P237
【参考文献】
相关期刊论文 前10条
1 徐岳仁;何宏林;申旭辉;陈立泽;;利用CBERS数据开展资兴市东江水库库区地质灾害动态监测试验研究[J];北京大学学报(自然科学版);2011年04期
2 刘秀华,刘勇,邵景安;重庆不同经济区土地利用与覆盖变化及社会驱动力研究[J];重庆大学学报(社会科学版);2003年02期
3 彭笃明;张安定;李德一;米姗姗;;基于LBV变换的遥感影像多步骤分类法研究[J];测绘科学;2008年03期
4 周启鸣;;多时相遥感影像变化检测综述[J];地理信息世界;2011年02期
5 申邵洪;赖祖龙;万幼川;;基于融合的高分辨率遥感影像变化检测[J];测绘通报;2009年03期
6 苗云鹏;张瑞红;赵保国;;高分辨率遥感影像信息提取方法研究[J];测绘与空间地理信息;2010年04期
7 张成雯;唐家奎;米素娟;胡引翠;王春磊;于新菊;李勇志;赵理君;王后茂;;中巴02B卫星多光谱影像中LBV数据变换方法研究[J];地理与地理信息科学;2011年03期
8 刘纪远,dls.iogac.on,布和敖斯尔;中国土地利用变化现代过程时空特征的研究──基于卫星遥感数据[J];第四纪研究;2000年03期
9 刘浩;胡卓玮;赵文慧;;基于面向对象的重大工程土地利用变化信息提取——以国家体育场(鸟巢)建设工程为例[J];国土资源遥感;2009年04期
10 陈宇;杜培军;唐伟成;柳思聪;;基于BJ-1小卫星遥感数据的矿区土地覆盖变化检测[J];国土资源遥感;2011年03期
，本文编号：2205103
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2205103.html
