標題: 三維雷射掃瞄儀點雲資料密度對最適平面求解精度影響
The Relationship Between Point Cloud Density and Fit-plane Accuracy of 3D Laser Scanners
作者: 徐元俊
Yuan-Chun Hsu
陳春盛
Chun-Sung Chen
土木工程學系
關鍵字: 三維雷射掃瞄儀;點雲密度;重新取樣;3D Laser Scanner;Point cloud density;Resampling
公開日期: 2005
摘要: 三維雷射掃瞄儀是一種能在短時間內快速獲取大量高精度三維點位相對坐標的儀器。它詳細地記錄了待測物體及其周遭景物的三維地表資訊。不同於以往測量儀器「點」的測量,而是以「面」的方式對周遭景物及待測物進行整體掃瞄。三維雷射掃瞄儀所得的初步資料是均勻密佈於被測物體表面的三維點雲資料,它仍是屬於隱性資料,對於被測物並非顯性描述,對大多數的工程測量應用而言,需要將點雲資料進行模型化的處理,以便獲得被測物之物件式或圖元式的顯性描述。
空間中的幾何資訊一般為物體的點、線、面,而三維雷射掃瞄儀所獲得的點雲資訊並無法直接獲得被測物表面的幾何資訊,雖然某些儀器廠商提供了原廠覘標以供捕捉特定點位的精確三維坐標,但是在實際外業工作上,除了佈設覘標的困難,如:高樓外側、建物邊角等;此外,具文化價值的傳統古蹟建物,也不容許有破壞性佈標工作的進行,因此如何利用既有的點雲資料來獲取被測物體的表面資訊也更顯得重要。
本研究所使用之點雲資料是以Trimble GS200型三維雷射掃瞄儀所得地面結構物表面之點雲資料。利用最小二乘法進行點雲最適平面的計算工作與精度分析,並利用不同的取樣方式,減少原始點雲群資料,分析其成果與原始點雲資料之差異,以期透過重新取樣的方式減少點雲資料,進而增進計算速度。
The 3D Laser Scanner is a kind of instrument that can obtain a large number of high-accuracy three-dimensional relative coordinates in a short time. The original point cloud data of objects are implicit not explicit. As to the applications to engineering, we have to modeling the point cloud data so as to get the explicit description of objects.
The geometry information of an object in the space are generally points, lines, and surfaces. These information can’t directly obtained from point clouds, though some instrument manufacturers have offered targets for catching the perfect three-dimensional coordinates, but in the field works, it is quite difficult. How to utilize the point clouds to obtain the geometry information of object surfaces is more important.
The test data applied in this research is scanned by Trimble GS200 3D Laser Scanner. Utilize Least-Square method to calculate the fitting planes and use different resampling methods to reduce the original point cloud data to understand the differences and improve calculate speed.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT009316578
http://hdl.handle.net/11536/78704
顯示於類別:畢業論文