基于高分遥感影像的建筑物轮廓规则化算法

Regularization of Building Outlines Based on High-Resolution Remote Sensing Images

针对高分遥感影像中提取的建筑物轮廓存在的残差和锯齿化问题,本文提出一种递进式轮廓规则化方法。在提取轮廓基础上利用改进的Harris检测算法初步实现建筑物轮廓边缘清晰化;然后基于Fréchet距离的最小面积外接矩形对初步边缘规则化的建筑物轮廓拟合优化,得到初步规则化轮廓;然而局部边缘转折区域仍存在锯齿状等问题,利用Shi-Tomasi算法对锯齿状局部区域进行优化,实现轮廓深度规则化。通过对比,本方法总精度达到了85.36%,相较于初始轮廓提高了13.17%,表明本方法适用于建筑物的外部轮廓线规则化,有效提高了建筑物外部轮廓线与边缘图形的表达率和精度。

Aiming at the problems of residual error and jagged edges exist in the building outlines extracted from high-resolution remote sensing images,a progressive outline regularization method has been proposed in the present study.On the basis of extracting the outline,the improved Harris detection algorithm is used to initially realize the edge of the building outline clearness;then the minimum area circumscribed rectangle based on the Fréchet distance is used to fit and optimize the building outline with the preliminary edge regularization,and the preliminary regularized contour is obtained.However,the problems such as jagged edges still exist in the local turning area.The Shi-Tomasi algorithm is used to optimize the jagged local area to realize the regularization of the outline depth.By comparison,the total accuracy of this method reaches 85.36%,which is 13.17%higher than the initial outline.It shows that this method is suitable for the regularization of the external contour lines of buildings,and effectively improves the graphic expression rate and accuracy of the external contour lines and edges of buildings.