本文選題：數(shù)字高程模型　切入點(diǎn)：離散點(diǎn)　出處：《山西農(nóng)業(yè)大學(xué)》2014年碩士論文

【摘要】：隨著計(jì)算機(jī)軟硬件技術(shù)發(fā)展,從上世紀(jì)50年代末至今,數(shù)字高程模型構(gòu)建算法一直是地學(xué)經(jīng)典研究領(lǐng)域之一,經(jīng)過諸多學(xué)者們的不斷努力,格網(wǎng)模型、不規(guī)則三角網(wǎng)模型、等高線模型等一系列成熟的模型應(yīng)運(yùn)而生,在模型構(gòu)建方面,也產(chǎn)生了一些比較成熟的思想。諸多數(shù)字高程模型(DEM)中以TIN較為常用,因此其生成的相關(guān)算法一直備受關(guān)注。 本文對(duì)傳統(tǒng)不規(guī)則三角網(wǎng)生成算法和基于DTIN追蹤生成等值線原理進(jìn)行概述,對(duì)它們主要特點(diǎn)進(jìn)行分析,結(jié)合傳統(tǒng)方法對(duì)整個(gè)過程進(jìn)行研究。算法實(shí)現(xiàn)涉及四個(gè)方面：離散點(diǎn)三角形剖分建立不規(guī)則三角網(wǎng)(D-TIN),基于D-TIN內(nèi)插生成等值點(diǎn)集合,開閉兩類等高線同時(shí)繪制,等值線適度光滑和標(biāo)注。關(guān)于離散點(diǎn)剖分生成D-TIN的過程,通過對(duì)傳統(tǒng)方法的對(duì)比,采用了一種相對(duì)實(shí)用的三角網(wǎng)生長(zhǎng)算法,它以三角網(wǎng)生成的一般算法為基礎(chǔ),對(duì)于生長(zhǎng)算法來說減少最鄰近點(diǎn)查找次數(shù)、降低計(jì)算復(fù)雜度是兩個(gè)主要的改進(jìn)方向,本生長(zhǎng)算法擴(kuò)展的內(nèi)容是多邊形,而不是三角形,但就三角形而言,每個(gè)三角形僅有兩條邊進(jìn)行拓展,而不是三條邊,在以基線為基礎(chǔ)的最鄰近點(diǎn)判斷上,進(jìn)行細(xì)化,將離散點(diǎn)以基線(有向線段)為分界分為左右兩個(gè)點(diǎn)集(剛好落在直線上的離散點(diǎn)與基線無法構(gòu)成三角形,所以不考慮),使用所需要的點(diǎn)集進(jìn)行判斷,一定程度減少了最鄰近點(diǎn)判斷次數(shù),只用判斷夾角是否最大,不用判斷是否位于左側(cè)(右側(cè)),以上是對(duì)一般算法的較大改良。值得一提的是在D-TIN數(shù)據(jù)結(jié)構(gòu)設(shè)計(jì)上,吸收傳統(tǒng)數(shù)據(jù)結(jié)構(gòu),以ArcEngine提供的PointClass、LineClass兩個(gè)類為基礎(chǔ)補(bǔ)充一些新屬性,分別自定義相應(yīng)的CoordinatePoint、Edge、Triangle類,提出新的儲(chǔ)存點(diǎn)、邊線的新結(jié)構(gòu),更加有利于等值點(diǎn)的計(jì)算和后續(xù)等高線的追蹤。傳統(tǒng)等高線追蹤過程是分開等高線和閉等高線分別進(jìn)行的,本文對(duì)等值點(diǎn)計(jì)算和等高線追蹤模式進(jìn)行了新的嘗試,首先進(jìn)行等值點(diǎn)的計(jì)算,將計(jì)算結(jié)果存儲(chǔ)到一個(gè)集合中,然后解析等值點(diǎn)集合,同時(shí)進(jìn)行開閉等高線追蹤(等高線采用ArcEngine提供的PolylineClass類對(duì)象進(jìn)行渲染)。采用PolylineClass類的Smooth函數(shù)對(duì)等高線進(jìn)行光滑,該方式簡(jiǎn)便快捷,可以直接將組成等高線的折線段集按照對(duì)應(yīng)參數(shù)轉(zhuǎn)換成相對(duì)平滑的貝塞爾曲線。最后使用ArcEngine提供的方法對(duì)等高線圖層進(jìn)行標(biāo)注。 經(jīng)過測(cè)試,離散點(diǎn)生成D-TIN的實(shí)用算法在一定程度上提高了構(gòu)網(wǎng)速度,具有很強(qiáng)實(shí)用性;同時(shí)也對(duì)新追蹤模式進(jìn)行驗(yàn)證,結(jié)果也是可靠的。
[Abstract]:With the development of computer software and hardware technology, from the late 1950s to the present, digital elevation model construction algorithm has been one of the classical geoscience research fields. Through the continuous efforts of many scholars, grid model, irregular triangular network model, A series of mature models, such as contour models, emerge as the times require, and some mature ideas have been produced in the construction of models. In many digital elevation models, TIN is commonly used, so the related algorithms have been paid more attention to. In this paper, the traditional algorithm of irregular triangulation generation and the principle of generating contours based on DTIN tracing are summarized, and their main characteristics are analyzed. Combined with the traditional method, the whole process is studied. The algorithm involves four aspects: establishing irregular triangular mesh by dividing discrete points into triangles, generating equivalent points set based on D-TIN interpolation, drawing two kinds of contour lines simultaneously. On the process of generating D-TIN by dividing discrete points, by comparing the traditional methods, a relatively practical algorithm of triangulation is adopted, which is based on the general algorithm of generating triangulation. For the growth algorithm, reducing the number of nearest neighbor lookups and reducing the computational complexity are the two main directions of improvement. The extension of this algorithm is polygon, not triangle, but for triangles, Each triangle has only two sides to expand, rather than three, to refine the judgment of the nearest point based on the baseline. The discrete points are divided into two sets of left and right points by the boundary of the baseline (the directed line segment). The discrete points that fall on the line and the baseline can not form a triangle, so the necessary set of points is used to judge the discrete points without considering the point set. To some extent, the number of nearest points is reduced, only the maximum angle is judged, and the left side is not determined (right side, above is a great improvement on the general algorithm. It is worth mentioning in the design of D-TIN data structure. Absorbing the traditional data structure and adding some new attributes based on the two classes of Point Class LineClass provided by ArcEngine, the corresponding coordinate Point Edge-Triangle class is defined, and a new storage point and a new structure of edge line are proposed. The traditional contour tracing process is carried out separately from the contour line and the closed contour line. In this paper, a new attempt has been made to calculate the equivalent point and track the contour line. First, the equivalent points are calculated, the results are stored in a set, and then the set of equivalent points is analyzed. At the same time, the contour is traced (the contour is rendered by PolylineClass object provided by ArcEngine. The contour is smoothed by the Smooth function of PolylineClass class. This method is simple and fast. The curve set composed of contour line can be directly converted into a relatively smooth Bezier curve according to the corresponding parameters. Finally, the contour layer can be annotated by using the method provided by ArcEngine. The test results show that the practical algorithm for generating D-TIN from discrete points improves the speed of network construction to a certain extent and has strong practicability. The new tracing mode is also verified and the results are reliable.
【學(xué)位授予單位】：山西農(nóng)業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：P208

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 韓光瞬;郭金麗;;顧及特殊地物的等高線自動(dòng)生成淺析[J];北京測(cè)繪;2009年04期

2 蔣詩洋;謝靈斌;;數(shù)字化測(cè)圖軟件中等高線處理初探[J];地礦測(cè)繪;2007年04期

3 彭波;鐘仕全;蘇旭芳;李雪;;漓江流域降雨量空間插值精度分析[J];氣象研究與應(yīng)用;2011年04期

，

本文編號(hào)：1666000