《基于遥感的农业用地变化检测技术研究》-毕业论文（设计）.doc

PAGE 河北工业大学2014届本科毕业设计论文 毕业设计（论文）中文摘要 基于遥感的农业用地变化检测技术研究 摘要： 遥感技术是近年来随着航天遥感技术的HYPERLINK /fazhan/发展而兴起的一种探测技术，高分辨率图像的出现，使提取农业土地利用变化信息变得更加有效而快捷。本课题以Landsat多光谱数据为研究对象，结合相应的指数，利用压缩数据维的方法进行分类后变化检测，获取研究区域主要土地利用类型的变化信息。对所要研究区域的地类以及地物分布特点进行分析，选取不同的指数组合来表示实验区的三大土地利用类型，即植被、水体及建筑用地。分析TM影像三个指数波段的相关性；选择分类样本；对指数影像进行分类后变化检测。试验结果表明指数模型不仅降低了数据维数，还减少了不同地类的光谱混淆；而改进的指数模型优于原始的指数模型。改进后的指数具有更好的分类精度，便于提取更加准确的变化信息。 关键词： 遥感技术 数据维压缩 土地利用 变化监测 监督分类 2014届本科毕业设计论文 毕业设计（论文）外文摘要 Title Agricultural land use change detection technology based on Remote Sensing Abstract With the development of space remote sensing technology,remote sensing technology is a detection technology risen in recent years,the advent of high resolution image makes extracting the land use change information become more effective and efficient. This paper takes Landsat multispectral data as the object of study, combines with the corresponding index,and detects the change after carrying out classification by using the method of data dimension compression . We select different index models to represent the three major land use types in experimentation area, through analyzing the land use types and characteristics of feature distribution in the study region .namely vegetation, water and land for construction. Analyze correlation of three index bands; select classifying instances;then detect the changes of the index image after Classification.The test results show that the exponential model not only reduces the dimension of data, but also reduces the spectral confusion among different kinds of land; and the improved index model is better than the original index model.The classification accuracy of the improved index model is better and easy to extract more accurate change information. Keywords： remote sensing technique data dimension compression land use change monitoring Supervised classification 目 录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc26020 1 绪论 PAGEREF _Toc26020 1 HYPERLINK \l _Toc21938 1.1 遥感技术 PA