2001-08-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/704880摘要：攝影幾何的求解一直是攝影測量流程中，極為重要的一環。唯有確定攝影幾何，建立像空間、物空間之函數關係後，利用像片或影像萃取物空間三維資訊才為可能。近十年來數位攝影測量藉助於影像處理技術及電腦視覺之判識能力、自動化特徵物偵測及匹配技術的改良， 許多不完整之特徵物資訊，例如點線面之幾何資料，隨著上述流程產生。究竟這些多樣式、多維的特徵物能為攝影測量作業帶來什麼樣的衝擊或幫助? 本計劃從影像特徵物的特性出發， 透過電腦計算、處理程序之分析，探討如何利用方便、豐富的特徵物解算攝影幾何；同時考量如何將影像中二維的特徵物轉化為三維物空間有意義之資訊。而最後階段效益之評估，旨在為未來的發展提供拙見。 <br> Abstract: The determination of imaging geometry, namely the orientation, has been a crucial step in photogrammetric workflow. Only when the knowledge about where and how the image was taken clarifies itself, does the extraction of object information photogrammetrically become easier or possible. On the other hand, the tremendous development of image processing technique as well as tools of computer vision showed significant impact on the progress of digital photogrammetric during the last decade. The most exciting observation is that, among others, the feature extraction and feature corresponding (or matching) algorithms are likely to detect and find many useful, though incomplete, feature information, such as point, linear and planar type features. Well, it’s not the end of the story. The challenge arises when one really tries to do something meaningful by employing these versatile features. For what ? and how? It’s the motivation of this study for taking care of the above issues from photogrammetric point of view. The central focus will be put on the characteristics of the features followed by how imaging geometry is resolved by way of the analyses and computations of features through the aid of the computer. In the meantime, the way 3-D object information resulting from 2-D image features reveals another core topic for this study. It’s author’s greatest hope that the evaluation of this work during the final stage will pave the right way for many follow-up research to be around.攝影幾何特徵物偵測影像處理匹配電腦視覺Imaging geometryFeature extractionImage processingCorrespondenceComputer vision