【摘要】 基于高分辨率多视图卫星图像，设计了一种用于生成时序三维建筑地图的方法.该方法采用基于对象的图像到地图变化检测方法来更新建筑物足迹，不仅用光谱和结构特征来提取非建筑物和建筑，而且还通过多视图图像的半全局匹配来自动估计建筑物高度.在实验评估方面，使用2013年至2018年的地图数据来验证本方法的有效性.由实验结果可知，本算法在估计建筑物足迹和高度方面具有较高的准确性.更多还原

【Abstract】 Based on high-resolution multi-view satellite images, a method for generating sequential 3 D architectural map is designed. This method uses the object-based image to map change detection method to update the building footprint. It not only uses the spectral and structural features to extract non-buildings and buildings, but also automatically estimates the building height through the semi-global matching of multi view images. In terms of experimental evaluation, the map data from 2013 to 2018 are used to verify the effectiveness of this method. The experimental results show that the algorithm has high accuracy in estimating the footprint and height of buildings.更多还原