本文關(guān)鍵詞：基于NewMap World的傾斜攝影數(shù)據(jù)三維可視化表達(dá)及單體化的研究與實(shí)現(xiàn)　出處：《太原理工大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 三角面片 求交檢測(cè) 分離 切割 單體化

【摘要】：單體化的對(duì)象是三維模型,目的是對(duì)某個(gè)模型進(jìn)行單獨(dú)管理與分析。傾斜攝影單體化即將需進(jìn)行單獨(dú)分析的三維模型與其它的三維模型分離開來,使得其可進(jìn)行單獨(dú)的管理和分析,而不是僅僅被用來查看。傾斜模型是專業(yè)建模工具處理傾斜三維數(shù)據(jù)后的產(chǎn)品,整個(gè)過程是自動(dòng)執(zhí)行的。建模過程中的源數(shù)據(jù),即傾斜攝影影像,首先根據(jù)源數(shù)據(jù)生成稠密點(diǎn)云,由于稠密點(diǎn)云不利于三角網(wǎng)的構(gòu)建,應(yīng)先抽稀再進(jìn)行構(gòu)建,最后貼上紋理貼圖,在這個(gè)過程中沒有人為干預(yù)。現(xiàn)有的建模算法并不能夠?qū)?shù)據(jù)中不同類型的三維模型分離,整個(gè)三維數(shù)據(jù)一整片顯示,數(shù)據(jù)層是連續(xù)不規(guī)則的三角網(wǎng)。若不能區(qū)分建筑、地面、樹木等地物,這樣的數(shù)據(jù)自身無法對(duì)單個(gè)建筑分離或選中,需要進(jìn)一步的處理才可進(jìn)行單體化。社會(huì)高速發(fā)展,用戶要求也越來越高。三維地理信息系統(tǒng)深入民眾生活,也期望其擁有新功能,能夠帶來更大便利。因此,它也不再局限于給人華麗的三維效果,或僅僅用來模擬�，F(xiàn)在,對(duì)于大多數(shù)的三維應(yīng)用而言,需要對(duì)其中的三維模型進(jìn)行單獨(dú)管理和分析,并對(duì)其中的模型賦予屬性,由系統(tǒng)提供查詢屬性的功能。無論是賦予屬性或是進(jìn)行查詢,都需要以單體化為基礎(chǔ)。因此,單體化成為三維應(yīng)用中必定要處理的問題。根據(jù)傾斜攝影單體化原理,單體化可以分為切割單體化和非切割單體化。非切割單體化只在視覺效果給觀察者不一樣的體驗(yàn),并沒有進(jìn)行單體化分離,不是徹底的單體化。切割單體化的優(yōu)點(diǎn)是實(shí)現(xiàn)了單體化分離,但已有的切割單體化方法基于點(diǎn)集,單體化模型邊緣效果較差。本文采用切割三角面片的方式進(jìn)行單體化,對(duì)目標(biāo)三維模型進(jìn)行邊緣處理,保證單體化三維模型的邊緣效果,使模型邊沿與屏幕分辨率保持統(tǒng)一。本文將應(yīng)用三角面片的可分割特性,通過切割分離三角面片的方式解決傾斜攝影數(shù)據(jù)三維模型可分離單體化問題。該方法首先繪制切割多邊形,確定單體化模型邊界。然后應(yīng)用包圍盒,過濾掉不可能與切割多邊形投影相交的三角面片,從而減少求交點(diǎn)時(shí)的實(shí)際運(yùn)算量。同時(shí)獲得與切割多邊形投影相交可能性較高的三角面片,計(jì)算這些三角面片與切割多邊形投影的邊的交點(diǎn)。多邊形邊與三角面片的相交情形有很多種,需要進(jìn)行詳細(xì)的分類分析,并針對(duì)不同的相交情形采用不同的切割方法,對(duì)相交區(qū)域三角面片裁切并重構(gòu)。三角面片做了裁切,其對(duì)應(yīng)的紋理也應(yīng)裁切,再通過紋理重構(gòu),完成貼圖。實(shí)驗(yàn)結(jié)果表明,本方法可有效實(shí)現(xiàn)傾斜攝影三維模型的分離單體化。
[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.

【學(xué)位授予單位】：太原理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 白利芳;常朝穩(wěn);王禹同;倪f，

本文編號(hào)：1401879