基于分类与形态综合的高分辨率影像建筑物提取方法研究

本文提出了一种基于分类和形态综合的高分辨率影像建筑物提取方法。主要研究的是基于高分辨率影像多尺度分割的面向对象人工地物分类方法,并以此为基础提取房屋,实现房屋位置、范围的初步自动化提取。最后,对提取的地物图斑的精确边界利用Hough变换提取直线边缘轮廓线,并以此为基础进行轮廓拟合和制图综合,从而得到规则的、方形组合的建筑物形状信息。

Method of building extraction from high resolution image based on classification and shape generalization

In this paper, the authors develop a new building extraction system applied on high resolution remote sensing image based on classification and shape generalization. Then they use multi-scale segmentation and object oriented classification to classify the high resolution multispectral image and extraction building areas. For shape reconfiguration of the building polygons, they adopt the probabilistic Hough transform to delineate the roof dominant line which shows the major orientation of the specific building roof. Based on their study, they perform the contour fitting and cartographic synthesis. Experiment shows that most rectangular building roofs can be correctly detected.