利用倾斜摄影测量的农村房屋信息调查

Rural Housing Information Survey Based on Oblique Photogrammetry

针对传统房屋测量投入周期长、测绘成本高的问题,提出了一套从数据获取到产品输出的农村房屋信息调查自动化解决方法。首先,利用无人机装载多视角相机以获取倾斜摄影测量影像数据,并从中生成三维点云、数字地表模型、正射影像图;然后,采用颜色与尺度信息融合的方式去除非建筑物点,利用多规则区域增长算法提取建筑物点云;最后,利用房屋外墙投影线、高程急剧变化边界线、屋顶不同平面相交线融合修正得到精确的房屋轮廓,并进行房屋空间信息计算与三维重建,实现了高效率、低成本、高精度的农房信息提取。在武汉新洲区施庙村的试点研究结果表明,自动化调查方法房屋面积在限差以内的房屋可达80%以上,平均精确率达到99.24%。

According to the fact that there is long investment period and high cost of surveying in traditional house surveying,the research presents a fully automatic approach for rural housing information survey from data acquisition to product output based on UAV tilt photography.Firstly,use the tilt photography data obtained by multi-directional camera mounted on the UAV to generate three-dimensional point cloud,digital surface model and digital orthophoto Map.Then,the method of combining color and scale information is used to remove non-building points,and multi-regular area growth algorithm is used to extract building points.Finally,use the projection line of exterior wall,the boundary line with sharp elevation change and the intersection line of different roof planes to fuse and correct the accurate house contour,carry out the calculation and 3D reconstruction of the house spatial information,and realize the high efficiency,low cost and high precision agricultural house information extraction.Our results of pilot study in Shimiao village,Xinzhou district,Wuhan indicate that the area within the limit of the housing proportion can achieve 80%,and the average precision rate is 99.24%.