行動測量系統可快速且有效地獲取空間的三維資訊，並且可提供細緻與高精度的資料。現今都市快速的發展以及電腦軟硬體的提升，在都市更新之規劃與維護、車輛導航、適地性服務等各個領域對於道路資訊的需求也逐漸增加，有效率的獲取細緻道路資訊為一重要之課題。本研究之目的為透過車載光達獲取的點雲資料中，萃取道路點雲，並從中獲取道路的幾何資訊。研究方法包含資料前處理、車道點萃取及驗證分析。因車載光達提供的資料量龐大，為提升運算效能，資料前處理以車載光達的軌跡計算道路中心線，並沿著中心線方向將資料切割成較小的路塊，分批進行處理。車道點萃取的部分，首先搜尋路塊中的高程最低點，並以該點為基準，設定門檻值保留可能為車道的點雲。接著，以三次多項式擬合道路平面，完成車道點雲的萃取。驗證分析的部分，以地測資料進行平面精度驗證的，高程以人工編修後的DEM為依據，驗證成果的正確性。

Mobile Mapping System (MMS) is a cost-effective way to acquire spatial data. It also provides detailed and high accurate point cloud. As the need for urban planning, navigation and Location Based Service (LBS) are increasing, how to extract the information of the road becomes the important research. For this reason, the purpose of this study is extracting the road points from the mobile lidar points cloud, and then obtaining the geometrical information of the road. The research has three major parts. There are data pre-processing, road point extraction and verification. In data pre-processing, in order to avoid dealing with all data, point clouds are divided into several road parts. Then, we select the lowest point as potential ground point from all points using elevation threshold. Next, we use the cubic curve fitting and point-to-curve distance to extract road points. It can remove non-ground points like cars and pedestrians. Finally, we provide manual edited digital elevation model (DEM) and ground check points to verify the result.