空載雷射掃描為一可快速獲取地面物體三維空間資訊之技術，而新型發展之全波形（Full-Waveform）系統可完整記錄雷射回波訊號之波形，透過波形偵測與波形擬合等資料前處理，可得到代表地物獨特反射特性的波形參數資料，包括振幅值（Amplitude）、波形寬（Pulse-width）與後續計算之散射截面積係數（Backscatter cross-section coefficient）。 得到各點位之波形資料後，將以波形資料為主進行位於山區之實驗區地物分類，並將使用由實驗區航照影像提供之RGB波段光譜資料計算之綠度指數（Greenness）與計算影像灰階統計值之紋理參數如均質度（Homogeneity）、熵值（Entropy）與R波段平均值（Mean）等參數輔助分類。分類進行之前，透過抽樣實驗區候選地類包括樹林、草地、道路與樹種建物，並以貝氏定理（Bayes Theorem）分析計算不同地物類別在各分類參數區間內的貝氏機率，接著以多項式函數擬合各地類在不同參數之貝氏機率曲線，並以計算反曲點之方式自動化決定該分類參數之門檻值區間。 分類成果顯示，全波形系統提供之波形資料對於受上層植物遮蔽與陰影區之植物點與道路點之分類有顯著之成果，且透過物體對於波形資料之反射特性不同，具備應用於區別不同建築材質類別之潛力。 / Airborne Laser Scanning is a technique capable of acquiring 3D information of land objects. The latest full-waveform system is further improved with the ability of recording complete waveform of reflected laser signal. After the preprocessing procedures such as pulse detection and pulse fitting, the waveform information including amplitude, pulse width and backscatter cross-section were derived. Such information was valuable as they represented unique properties of land objects. In this study, waveform information of all scanned points were utilized to classify land cover in the test area located in mountain area. Additionally, the Greenness value as well as the texture parameters such as Homogeneity, Entropy and Mean of R band calculated from the ortho-image were used for classification. We aimed to classify the point cloud into vegetation, road and building categories. The Bayes Theorem was used to determine the threshold range of each parameters for classification. As a result, the waveform information were useful for classifying road points covered by upper vegetation points and also vegetation and road points located in shadow area. Moreover, through the analysis of reflective properties of different object using waveform parameters, it was of potential to be applied to distinguish material of buildings.