模型导向的建筑物半自动三维重建

发布时间：2018-05-26 23:13

  本文选题：模型导向 + 建筑物三维建模　； 参考：《遥感信息》2017年03期

【摘要】：针对以航空影像为数据源的建筑物三维重建中存在的自动化程度低、精度不高、计算量大的问题,提出一种针对任意建筑物的半自动建筑物三维建模方法。对传统的模型导向的方法进行改进,首先建立基本模型库,从模型库中选择模型组建建筑物,根据建筑物内部约束确定建筑物参数;然后提出依据影像上提取的建筑物边缘信息计算建筑物参数初始值的方法;最后使用广义点摄影测量理论进行平差运算计算建筑物的精确参数。针对几种不同类型的建筑物进行实验,结果表明,该方法切实可行,可以对不同形状的建筑物重建,且具有精度高、计算速度快、人工辅助少的优点。
[Abstract]:In order to solve the problems of low automation, low precision and large amount of calculation in building 3D reconstruction based on aerial image, a semi-automatic building 3D modeling method for arbitrary buildings is proposed. The traditional model-oriented method is improved. Firstly, the basic model base is established, the model is selected from the model base to build the building, and the building parameters are determined according to the internal constraints of the building. Then a method is proposed to calculate the initial values of building parameters based on the building edge information extracted from the image, and the accurate parameters of the building are calculated by using the generalized point photogrammetry theory. Experiments on several different types of buildings show that this method is feasible and can be used to reconstruct buildings with different shapes, and has the advantages of high accuracy, fast calculation speed and less artificial assistance.
【作者单位】： 河南大学环境与规划学院;
【基金】：国家自然科学基金(41101450) 河南大学优秀青年科研人才培育基金项目(yqpy20140016)
【分类号】：P231
，

本文编号：1939338