摘要
倾斜影像生成的城市场景三维模型是一个整体的DSM,然而城市管理中需要单体且部件化的精细三维模型,依据影像或者密集匹配点云进行建筑物的自动分层是实现精细化建模的重要环节。文中以倾斜影像构建的场景模型中的单体建筑物三维模型为分析对象,生成其不同方向的立面纹理影像,将各立面纹理影像转换为累加梯度信号,采用一维离散小波变换(DWT)方法提取出楼层的分层位置,并依此对多层建筑物进行楼层划分,最后根据单层点云数据构建单层网格模型。文中方法的结果表明,建筑物楼层划分误差为0.05 m,构建的单层模型为下一步分户数据的提取及模型的构建做准备。
The 3D model of urban scene generated by tilt photogrammetry is a whole DSM,however,the urban management application needs a single 3D model with different parts,and the automatic layering of buildings based on images or dense matching point clouds is an important part to reconstruct modeling.In this paper,we take a single 3D model of a building in a scene model constructed by tilted photogrammetry as the analysis object,generate its elevation texture images in different directions,convert each elevation texture image into a cumulative gradient signal,extract the layering position of floors by using the one-dimensional discrete wavelet transform(DWT)method,and then classify the floors of multi-layer buildings accordingly,finally construct a single-layer grid model based on the point cloud data.The results of the method in this paper show that the error of building floor division is 0.05m,and the single-layer model is prepared for the next step of extraction of subdivision data and construction of the model.
作者
肖洪瑞
盛业华
XIAO Hongrui;SHENG Yehua(School of Geography,Nanjing Normal University,Nanjing 210023,China;Key Laboratory of Virtual Geographic Environment of Ministry of Education,Nanjing Normal University,Nanjing 210023,China;Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application,Nanjing 210023,China)
出处
《测绘工程》
2023年第4期37-43,共7页
Engineering of Surveying and Mapping
基金
国家自然科学基金资助项目(41631175,42071364)。
关键词
倾斜模型
立面影像
楼层分割
小波变换
纹理
tilt model
facade image
floor segmentation
wavelet transform
texture
