森林資源的調查是森林管理與保護的一項重要工作。傳統上森林資源調查主要是依賴繁重且費力的地面調查。雖然已經有許多先進的儀器被用來改善地面調查的效率，但是仍有許多的研究利用遙測影像，以發展自動化的森林調查方法，特別是在立木的偵測及樹冠的描繪。一般影像上的立木型態，除了單一立木外，仍有部份立木因間距過度緊密，而造成樹冠範圍相鄰並且群聚成的樹叢。因此，本研究將偵測及描繪影像上的單一立木與樹叢中立木的個數及範圍。 本研究的目的主要是利用高解析的航照影像，發展一套以形態學(morphology)為基礎的演算法，以進行立木的偵測與樹冠之描繪。本研究方法可分為三大步驟: (1)樹物件的萃取，(2)多尺度立木結構元 (Structure Element, SE)的建構，以及(3)立木偵測及樹冠描繪。 本研究主要是使用高解析數位航照影像作為測試資料，測試影像上的立木主要為闊葉樹種，其分布型態包含尺寸不均勻的單一立木，及由數個尺寸不均勻的立木所組成的樹叢。本研究根據闊葉樹種的樹冠輪廓在影像上近似圓形的特性，偵測影像上不同尺寸的單一立木及樹叢中不同尺寸的立木個數，並描繪其範圍。本研究測試成果顯示，立木偵測的精確度約為98.8%，樹冠描繪的精確度約為92.9%。因此本研究之方法可以有效的偵測出影像上不同尺寸的單一立木及樹叢中不同尺寸的立木之個數，並描繪其樹冠範圍。 The survey of forest resources is an important task for the management and protection of the forest. Traditionally, such an essential task is heavily dependent on the labor-intensive ground survey. Although numerous state-of-the-art equipments have been developed to improve the efficiency of the ground survey, a lot of researches have been using remote sensing images for the forest survey, especially in the domain of the automatic detection of individual tree and the delineation of tree crown. The type of tree in the image not only contains individual tree, but also contains tree clump caused by over short distance between trees. Therefore, the purpose of this study is to detect and delineate the individual tree and tree clump in the image. The purpose of this study is to develop a new approach to detect the individual tree and to delineate the tree crown from high spatial resolution aerial photography. The method is based on the concept of mathematical morphology to perform the tree detection and delineation. The algorithm can be dividing into three steps: (1) the extraction of the tree objects, (2) the construction of the multi-size tree structure elements for the morphological operation, and (3) the accomplishment of the tree detection and delineation. The test data for this study are high spatial resolution digital aerial photography, and the species of trees on the test data are Hardwood. The distribution type of trees on the test data contain of multi-size individual trees and tree clumps with multi-size trees. The accuracy of the detection is approximately 98.8%, and the accuracy of delineation is approximately 92.9%. In this study, multi-size individual trees and the tree of the tree clumps can be detected effectively, and the tree crown also could be delineated accurately.
