基于几何规则性约束的建筑物点云张量投票提取方法的研究

论文总字数：20873字

目 录

摘要....................................................4

Abstract ................................................ 5

1 绪论................................................ 6

1.1 研究背景和意义................................................. 6

1.2 国内外研究现状................................................. 6

1.3 研究内容....................................................... 7

2 建筑物屋顶点提取.................................... 7

2.1 原始数据滤波处理............................................... 7

2.2 几何规则性约束................................................. 8

2.2.1 点密度约束................................................. 8

2.2.2 规则性约束................................................. 8

2.3 张量投票的方法提取建筑物屋顶................................... 9

2.3.1 传统张量投票阶段........................................... 10

2.3.2 传统张量分解阶段........................................... 10

2.3.3 投票域..................................................... 11

2.3.3.1 棒投票域............................................... 11

2.3.3.2 板投票域和球投票域..................................... 12

2.3.4 解析张量投票新机制进行建筑物屋顶提取....................... 13

2.3.4.1 棒张量投票解析解的求取................................ 13

2.3.4.2 板张量投票和球张量投票解析解的求取.................... 14

2.4 基于几何规则性约束的建筑物点云张量投票提取.................... 15

3 实验与分析.......................................... 15

3.1 原始数据的滤波处理............................................ 16

3.2 单一的张量投票建筑物屋顶提取.................................. 18

3.3 基于几何规则性的建筑物屋顶点云张量投票提取.................... 19

3.4 方法的比较和分析.............................................. 21

4 结论和展望......................................... 23

5 参考文献............................................24

6 致谢................................................26

基于几何规则性约束的建筑物点云张量投票提取方法的研究

吕来东

, China

Abstract:With the progress of The Times and the development of science and technology, the city we live in is undergoing earth-shaking changes. Urban digitization has gradually become the trend of The Times, while buildings in cities are the important elements of urban digitization and three-dimensional modeling, and the extraction of the roof of buildings is a key step in 3d modeling. At the same time, buildings are also important infrastructure of cities. The acquisition of boundary information is of great significance for land use planning, map updating and urban change detection. How to extract buildings from remote sensing data quickly and accurately has become one of the current research hotspots.

In recent years, laser LIDAR point cloud technology is becoming more and more mature, which also provides a large number of three-dimensional point cloud coordinate data for the research of buildings. This paper mainly USES the original laser LIDAR point cloud data (i.e., 3d coordinate information only) to realize the extraction of the roof of the building. The methods are as follows:

First of all, the original point cloud filtering processing, select the point on the ground, according to the rules of point cloud of geometric constraint, distinguish between what is about building points, which are the tall vegetation, which is roughly the extraction of building roof.

Secondly, by using the method of tensor voting, we can accurately distinguish between building points and vegetation points.In this case, I adopted a new tensor voting mechanism proposed by yanchuanShao. Compared with the traditional tensor voting, no integration iteration is required, and the experimental results are more accurate.

Thirdly, based on the advantages and disadvantages of the above two methods, an extraction method of building point cloud tensor voting based on geometric regularity constraint is proposed.

Finally, this method as mentioned here makes it possible to extract the roof cloud data of a building more accurately than the previous methods of extracting the roof of a building.

Keywords:Laser point cloud ; Point density ; Tensor voting ; Extraction of building roof

1 绪论

1.1 研究背景和意义

数字地球作为一个完整的名词，最初于1997年下半年出现在科学界。1998年1月31日，美国副总统戈尔在加利福尼亚科学中心作了题为“数字地球--认识二十一世纪我们这个星球”的讲话，率先在公开场合喊出“数字地球”这一名词。地球是由一座座城市组成的，也正是基于此，有人提出了“数字城市”这个概念，随着科技的进步和时代的发展，城市数字化正逐渐被越来越多的国家所重视。

改革开放经过了四十多年，我国经济飞速发展，城市化进程也正在有条不紊的进行中，因此，我国对于城市信息获取的要求也越来越高，城市数字化正成为一个主流的发展方向。

剩余内容已隐藏，请支付后下载全文，论文总字数：20873字

相关图片展示：