基于NewMap World的倾斜摄影数据三维可视化表达及单体化的研究与实现

发布时间：2018-01-09 15:10

本文关键词：基于NewMap World的倾斜摄影数据三维可视化表达及单体化的研究与实现　出处：《太原理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：单体化的对象是三维模型,目的是对某个模型进行单独管理与分析。倾斜摄影单体化即将需进行单独分析的三维模型与其它的三维模型分离开来,使得其可进行单独的管理和分析,而不是仅仅被用来查看。倾斜模型是专业建模工具处理倾斜三维数据后的产品,整个过程是自动执行的。建模过程中的源数据,即倾斜摄影影像,首先根据源数据生成稠密点云,由于稠密点云不利于三角网的构建,应先抽稀再进行构建,最后贴上纹理贴图,在这个过程中没有人为干预。现有的建模算法并不能够将数据中不同类型的三维模型分离,整个三维数据一整片显示,数据层是连续不规则的三角网。若不能区分建筑、地面、树木等地物,这样的数据自身无法对单个建筑分离或选中,需要进一步的处理才可进行单体化。社会高速发展,用户要求也越来越高。三维地理信息系统深入民众生活,也期望其拥有新功能,能够带来更大便利。因此,它也不再局限于给人华丽的三维效果,或仅仅用来模拟。现在,对于大多数的三维应用而言,需要对其中的三维模型进行单独管理和分析,并对其中的模型赋予属性,由系统提供查询属性的功能。无论是赋予属性或是进行查询,都需要以单体化为基础。因此,单体化成为三维应用中必定要处理的问题。根据倾斜摄影单体化原理,单体化可以分为切割单体化和非切割单体化。非切割单体化只在视觉效果给观察者不一样的体验,并没有进行单体化分离,不是彻底的单体化。切割单体化的优点是实现了单体化分离,但已有的切割单体化方法基于点集,单体化模型边缘效果较差。本文采用切割三角面片的方式进行单体化,对目标三维模型进行边缘处理,保证单体化三维模型的边缘效果,使模型边沿与屏幕分辨率保持统一。本文将应用三角面片的可分割特性,通过切割分离三角面片的方式解决倾斜摄影数据三维模型可分离单体化问题。该方法首先绘制切割多边形,确定单体化模型边界。然后应用包围盒,过滤掉不可能与切割多边形投影相交的三角面片,从而减少求交点时的实际运算量。同时获得与切割多边形投影相交可能性较高的三角面片,计算这些三角面片与切割多边形投影的边的交点。多边形边与三角面片的相交情形有很多种,需要进行详细的分类分析,并针对不同的相交情形采用不同的切割方法,对相交区域三角面片裁切并重构。三角面片做了裁切,其对应的纹理也应裁切,再通过纹理重构,完成贴图。实验结果表明,本方法可有效实现倾斜摄影三维模型的分离单体化。
[Abstract]:The object is a single 3D model, is to separate management and analysis of a model. The three-dimensional model of single tilt photography is to carry out the analysis of the 3D model and the other is separated, so that it can be managed and analyzed separately, rather than just being used to view the tilt model is professional. Modeling tool tilt 3D data of the product, the whole process is performed automatically. The source data in the process of modeling, namely tilt photographic image, according to the source data to generate a dense point cloud, because the construction of dense point cloud triangulation is not conducive to the first thinning and construction, finally paste texture, no human intervention in the process of modeling. The existing algorithms can not separate 3D models for different types of data, the entire 3D data display, data layer is continuous and irregular triangle net. If not Distinguish between buildings, ground, trees and other objects, so the data itself to a single building or separation is selected, need further treatment can be a single. The rapid development of society, users are increasingly high requirements. The three-dimensional geographic information system into people's lives, it is also expected to have new features, can bring greater convenience. Therefore, it is not only to give people the gorgeous 3D effect, or just used for the simulation. Now, for most 3D applications, the need for separate management of 3D model and the analysis, and gives the attributes of model, attribute query function provided by the system. It is given the attribute or query needs, based on the monomer. Therefore, monomer become bound to deal with problems in 3D applications. According to the principle of oblique photography of monomer, monomer can be divided into single cutting and non cutting Monomer. Non cutting monomer only in visual effect to the observer is not the same experience, and no single separation, not completely monomer. The advantages of cutting monomer is the monomer monomer separation, but cutting method based on point set, monomer the model of poor edge effect. This paper uses the triangle cutting way of monomer, the three-dimensional model of the target edge processing, ensure the monomer of the three-dimensional model of the edge effect, make the model with the edge of the screen resolution to maintain unity. This paper will use the integral characteristics of the triangle, the triangle cutting separation way tilt the 3D model of photography data can be isolated problem. This method firstly draw cutting polygon, determine the single model boundary. Then the application of bounding box, and filter out impossible triangle surface cutting polygon projection from the intersection To reduce the actual computation. At the same time when the intersection point and cutting polygon intersection projection higher possibility of triangles, the triangle intersection calculation and cutting edge. The projection polygon polygon edges and triangles intersect situation there are many, the need for detailed analysis and classification, and according to the different situations by using intersection cutting method different, on the intersection of triangles cutting and reconstruction. Triangles do cut, the corresponding texture should be cut, and then complete the texture by texture reconstruction. The experimental results show that this method can effectively realize the separation of single 3D model photography tilt.

【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】