基于基尼系数的倾斜航空影像中建筑物立面损毁检测

Building Facade Damage Detection Based on the Gini Index from Oblique Aerial Images

针对已分割出的建筑物立面,提出了以基尼系数作为损毁指数对倾斜航空影像中建筑物立面损毁检测的方法。该方法首先通过改进的k-means算法分割出建筑物立面的门窗和墙面,然后利用Canny算法对分割影像进行边缘检测提取统计特征,最后利用经济学中的基尼系数来判断建筑物立面是否损毁。实验结果表明,本文方法在没有先验信息的情况下仅利用单时相倾斜航空影像就能简单高效地判定建筑物立面损毁。

This paper proposes a building facade damage detection method using Gini index of economics from the building facade of oblique aerial images.This method firstly segments the windows and doors from the building wall using the improved k-means algorithm,then detects the edge using canny algorithm,at last judges whether the building facade is damaged using Gini index.Experimental results show that the proposed method can be used to detect the building facade damage simply and efficiently only using the single post-seismic oblique aerial images without any prior information.