高分辨率卫星遥感数据在土地利用动态监测中的应用研究

THE APPLICATION OF HIGH RESOLUTION SATELLITE REMOTELY SENSED DATA TO LANDUSE DYNAMIC MONITORING

选择土地利用变化类型较多的北京市大兴区为试验区 ,通过分析印度IRS、韩国KOMPSAT- 1、美国IKONOS等高分辨率卫星数据的获取途径、覆盖周期、卫星运行状况、监测成本估算、技术条件和设备要求 ,对其在土地利用动态监测中的可行性进行了研究 ;测算了高分辨率卫星数据的主要技术参数 ;分析原始图像上各主要土地类型的灰度分布范围及其直方图特征 ,总结不同土地利用覆盖类型与融合图像变化信息的光谱特征。通过精度测算 ,确定了最大成图比例尺。比较利用高分辩率卫星IRS、KOMPSAT - 1和SPOT与陆地卫星TM多光谱数据组合的土地利用动态监测结果 (识别率、判对率、图斑边界吻合精度、可监测的最小图斑面积等 ) ,确定了不同高分辨率卫星数据与TM多光谱卫星数据组合后对土地利用类型的识别能力 ,建立了土地利用动态监测解译标志。在解决了若干关键技术问题的基础上 ,形成了一套可推广的。

In this study, the authors chose Daxing County of Beijing as the experimental field in view of its varied landuse styles. Based on analyzing the data-acquisition ways, covering periods, operation conditions, costs, techniques and equipment conditions of some high resolution satellites such as Indian IRS, Korean KOMPSAT-1 and American IKONOS, the authors conducted researches on the feasibility of performing dynamic monitoring of landuse with remotely-sensed data, determined some important technical parameters in high resolution data, analyzed the range of gray level and characteristics of histogram for principal landuse styles, and summarized the spectral characteristics of different types of landuse and land cover. In addition, the spectral characteristics of merged images were analysed, the largest scale of mapping was determined, the monitoring results of different data sources such as IRS and TM, KOMPSAT-1 and TM as well as SPOT and TM were compared, the recognition capacities of different combinations of high resolution data and multi-spectral data were differentiated, the interpretation signatures of land usage dynamic monitoring were established, and some crucial technical problems were solved. A set of operational and practical technical methods and producing procedures have been set up, which serve as great technical supports to national land resources investigation.