隨著遙測科技的進步，空載光達所建構的樹冠高度模型（Canopy HeightModel, CHM）應用於樹冠孔隙與邊緣偵測之萃取技術日趨成熟。本研究以此概念建立阿里山地區之影像像元灰階亮度值的頻度分布圖（臺灣赤楊人工林、柳杉人工林、紅檜人工林與紅檜天然林），選取前兩個高峰間的波谷灰階亮度值作為門檻，並運用影像分割技術進行樹冠孔隙偵測，藉以建立樹冠孔隙空間分布。研究結果顯示，利用樹冠高度的觀念，以門檻值的設定方式獲取樹冠孔隙的空間分布，已證實可行；唯其所設定之門檻值，必須考慮氣候狀態與生育地環境，如樹冠高度、枝下高以及樣區孔隙現況資料。如在鳥隻與雲霧等所產生之雜訊影響不大的情況下，使用全區門檻值法，以區域二值化的處理，進行樹冠孔隙偵測，可達最佳效果，其總體精度為97.45％，Kappa值為0.74。

Following the progress of remote sensing technology, a Canopy Height Model (CHM) has commonly been applied for delineating canopy gaps automatically and extracting its edge detection by using an airborne LiDAR (Light Detection And Ranging). This study established the frequency distribution of gray-scale image pixel brightness values, selected the first two peaks troughs between the gray-scale brightness value as a threshold value, and used the image segmentation techniques to detect the boundary of crown and then established the distribution of forest gap. The results showed that it was feasible to obtain the spatial distribution of forest gap by using the concept of canopy height to set the threshold method. The only exception was that when they had set the threshold, it has to consider the characteristics of the physcial environments and climate conditions. For example, in the case of situating less-noisy environment the method of threshold value could show the best result by using the region to carry out binarization image processing to detect the crown boundary, because the Kappa value and overall accuracy of classification were increased to 0.74 and 97.45%, respectively.