土地利用遥感动态监测的技术及应用研究

【作者】 刘敬；

【导师】 李俊；

【作者基本信息】 昆明理工大学 ， 矿产普查与勘探， 2007， 硕士

【摘要】 土地利用动态监测是实现全面及时掌握资源状况的强有力的手段。传统的土地利用动态监测任务是由人工土地利用调查来完成的，但人工土地利用调查费时、费力、耗资、效率低下。随着遥感与地理信息系统技术的兴起与日益成熟，人们积极探索利用这些高科技成果，监测土地资源的利用及覆盖状况。本文在收集、总结国内外同类研究成果的基础上，采用遥感手段，以滇池流域(除嵩明外)为研究区域，2000年和2004年两个时相的陆地资源卫星(Landsat)TM和ETM为数据源，在专业遥感图像处理软件ERDAS IMAGINE8.7和MAPGIS6.5支持下，进行土地利用的遥感动态监测的技术及应用研究。在信息提取方面，对2000年ETM数据先采用小波变换进行数据融合，并对小波变换数据融合后影像的结果加以分析评价。在分类方法方面是结合合目视判读的经验，采用混合分类法。首先对两幅影像进行非监督分类，判断初始类别确认属性，将判读后的非监督分类的分类属性表经过光谱聚类处理转化成适用于监督分类的分类模板文件，然后再执行监督分类。两个时相影像的分类精度分别为2000年88％，2004年是82.5％，均达到分类的标准。根据分类成果进行研究区4年间土地利用数量结构变化分析、土地利用转型分析和土地利用驱动力分析。结果表明4年来该地区土地利用／土地覆被变化很大：建设用地面积以每年10.85％的速度增加，占用了大量的耕地；相应的耕地和未利用土地面积大量减少，减少率分别为46.23％和63.8％；水域和林草地面积虽有萎缩，但变化不大。从而分析得出影响土地利用变化内在驱动的主要原因为：工业化和城市化的快速发展、人口高速增长和城市规划对土地利用的导向性作用。更多还原

【Abstract】 Land utilization dynamic monitoring is to come true at present have forceful resource situation means in hand all- round , timely. The tradition land utilization dynamic monitoring mission is that land utilization is completed from the manpower coming investigation, but on inquiry, manpower land utilization wastes time and energy, blows the expense at, poor efficiency . Nowadays, with the remote sense and geographic information system technology being on the rise, with gradually, mature , fully utilizing, these high technology achievement, makes use of status and covers to monitoring land resources to become possibility and conduct institute common view.This dissertation discusses the techniques and methods regarding monitoring land Use dynamic county level based on change detection technique of remote sensing data Update and analysis based on software of ERDAS IMAGINE8.5 and MAPGIS6.5. The above methods and techniques are applied to the waters of DIAN CHI, land use changes from 2000 to 2004.On the extracted information aspect that transform the data fuse being in progress to ETM data on 2000 year first with wave, the result shifting the queen portrait’s gives analytical valuation to wave. Be to adopt to blend classification combining with closing visual be sentenced to read experience in the field of classification method. First, be in progress must super classification , judge that the initial category affirms an attribute to two portraits .second, The classification attribute mistake of the be sentenced to read queen is superintended classifying gauge becomes the classification form board document applying to supervision classification by that spectrum clustering handles conversion, and then carry out supervision classification again. Two time images, the portrait classification accuracy is that 88% in 2000, is 82.5% in 2004 , reaches the classed standard equally each other respectively. During the 4 years, land utilization quantity structure change analyses , land utilization revolution of type analysis and the land utilization driving force analyses the area being in progress according to classification achievement studying. Come result to be area land utilization/ land’s turn to be covered by change indicating 4 years very big: 10.85%’s speed has increased construction with field area with every year , has occupied large amount of cultivated land; Corresponding cultivated land and a great quantity notmaking use of land area fall off , the decreasing rate is 46.23% and 63.8% respectively; Though water area and forest grassplot area having, the atrophy, but change is not big. Analysis reaches the cause affecting inherent drive of land utilization change main part thereby being: The fleetness being industrialized and urbanizing develops、 high speed increases population and The urban planning has the guiding effect to land utilization.更多还原