Abstract: Automatic building extraction is the key factor of building map based on remote sensing image.Aiming at the limitations existing in the current algorithms，a novel algorithm for automatic building extraction based on Digital Elevation Model（DEM） is proposed.The algorithm，which takes the amount of calculation into account fully，is built on the base of mathematical grey scale morphology，vector conversion，discretization noise elimination and edge simplicity.Experimental results show that the algorithm can extract building clearly and correctly，and reduce the redundancy greatly.

Key words: automatic building extraction, Digital Elevation Model（DEM）, remote sensing image, image processing

摘要： 从遥感影像中自动提取建筑物是研究从遥感影像构建地图的关键技术。针对当前建筑物自动提取算法中存在的缺陷，提出了一种基于数字高程模型的建筑物自动提取算法。该算法是建立在灰度数学形态学、矢量化变换、离散化噪音消除以及边缘简化技术综合的基础上，充分考虑了建筑物后续重建计算量的问题。实验结果表明利用该算法提取出的建筑物边缘清晰正确，同时又能大大减少冗余数据量。

关键词: 建筑物提取, 数字高程模型, 遥感影像, 图像处理