城市高分辨率遥感影像房屋提取及变化监测.docVIP

本科毕业设计(论文)材料 GRADUATION DESIGN(THESIS) MATERIAL 题 目 城市高分辨率遥感影像 房屋提取与变化监测 学生姓名 王 兴 平 指导教师 陶 超 学 院 地球科学与信息物理学院 专业班级 测 绘 1103 班 2015年6月 城市高分辨率遥感影像房屋提取 与变化监测 摘要 当今，高分辨率遥感卫星已经从上世纪90年代的主要为军事服务，逐渐向产品商业化的趋势发展，许多基于高分辨率遥感影像的应用问题亟待研究。本文根据房屋的亮度、面积、形状以及各向同性的光谱特征与几何特征，运用顶帽变化和差分等数学形态学运算，对对多组不同时相的高分辨率遥感影像进行房屋提取，运用图像差分法与连通域重合度变化监测，并对发生变化的部分，提出了形态学边缘提取法与连通域角点法进行标记的方法。本文通过大量实验证明了形态学房屋指数（MBI,Morphological Building Index）法在高分辨遥感影像的房屋自动提取的可行性与精度，并且对比了多种变化监测标记方法各自的优缺点。高分辨率遥感影像信息自动提取的研究与应用，发展前景广泛。 关键词：高分辨率遥感影像 形态学房屋指数 房屋提取 变化监测 Building Extraction and Change Detection From High-Resolution Imagery Over Urban Areas Abstract Nowadays,the high-resolution satellite’s product has changed its main service field from the military to the commercial field since 1990s. Many applied issues based on high-resolution imagery need to be solved urgently. By considering the spectral feature and geometric feature, E.G. brightness, area and isotropic, this paper use the mathematical morphology algorithm such as top-hat transform and differential morphological profiles to extract buildings from a set of multi time high-resolution imageries, and detect the changes between these two different moments’ imageries of a same place by using the image differential method and the connecting the corners of a connected region method. This paper proposed an approach on marking the change areas on current high-resolution imagery. Based on large amount of experiment,it sated that it’s possible to use the MBI Algorithm to extract buildings, and promised a some accuracy, we contrast two different methods’ advantages and disadvantages. Research on getting information automatically from high-resolution imagery has deep developments. Key Words: High-resolution Remote Sensing, Morphological Building Index, Building Extraction, Change Detection. 目 录 第1章 绪论 1 1.1 课题研究的背景和意义 1 1.2 遥感技术