地面光達於崩塌地地形製圖之應用

Abstract

地面光達（Ground-based Lidar）是一種新穎的三維量測系統，其乃利用雷射測距之原理，在極短的時間內取得大量之被測體表面的點位三維座標，可詳細描述被測體之外型，更可進一步用於建物之三維模型建置或崩塌地等地形進行變形監測及調查。 傳統的測量技術僅能提供有限點位的資訊，而地面光達則能提供大量的點位資訊。另外，在儀器設計上，地面光達有許多不同於傳統測量儀器的特性，如不具有定心定平之設備，故資料結構上有其特殊設計以利於不同測站間之點雲資料疊合及座標轉換之用。 本研究中以竹東中豐路整治地及九份二山崩塌地為例，依序進行不同測站間點雲資料之連結、疊合及轉換，並由點雲資料中製作DEM資料，對數值影像進行幾何糾正後，將其與DEM資料套疊，用以製作三維透視圖。 從本研究成果可發現，透過地面光達所獲得之點雲資料可詳細地表現崩塌地之地勢起伏情形，可有效地用於崩塌地製圖。

Ground-based lidar is a modern 3D survey instrument, utilizing the principle of laser ranging to acquire a large amount of 3D points of an object in a very short time. The point clouds integrated from multiple stations can portray the object in details. Thus, it can be further applied to monitor the morphology and deformation of landslides. As compared to traditional survey methods which only offer coordinate information of limited points, ground-based lidar can offer massive three-dimensional information of surface shape. In addition, ground-based lidar has several unique features different from conventional total station type surveying instruments. For example, there are no centering and leveling device in most models of the current lidar systems. A specific design of data organization is required for the alignment of point clouds of multi-stations, geodetic registration and transformation. Two sites were included in this study, namely the Chu-dong river-side landslide and Jeofenerh-san large-scaled dip-slope landslide. A procedure for data processing to obtain a proper dataset for subsequent application was adopted. This includes the acquiring of the multiple scans of point clouds, alignment and merging, and geodetic transformation. Regular DEM is then interpolated from the point clouds for making contour maps and cross-sections of the landslides, and also for registering with digital oblique images of the area to further enhance the understanding of landslides. It is concluded in this study that ground-based lidar is an effective tool for landslide mapping which have the merits of fast generation of detailed 3D topographic information of the landslides. Therefore more supplying information than that of the traditional approaches.