影像視覺追蹤與定位技術可用於擴增實境之應用，依使用標記與否，可分為標記與無標記兩類技術。標記識別技術較為簡單，且較為成熟，但受限於需使用特殊之標記。無標記識別技術使用一般自然物件來取代特殊標記，應用較為彈性，但必須能正確擷取出欲辨識物件有用之特徵點。現有擴增實境之開發函式庫或開發平台，主要目的為提供擴增實境之應用開發，不提供無標記擴增實境技術研究方法之驗證。故本文深入研究探討如何整合ARToolKit、OpenCV與OpenGL等函式庫，設計無標記擴增實境整合平台。且此平台各項功能採用模組化設計，以更彈性提供適用於無標記擴增實境物件辨識與追蹤相關方法之驗證。另外，本文並探討自然物件複雜度與移動速度對辨識與追蹤之影響，分析適用於擴增實境自然物件之特徵值，做為後續提出克服因低複雜度自然物件以及高速移動導致追蹤失敗之缺失，藉以同時驗證平台之適用性。本文之研究結果，可做為實現開發無標記擴增實境實驗平台之重要參考依據。-Visual image tracking and positioning technology can be used to amplify the reality of the application, according to the use of tags or not, it can be divided into two types of markers and unmarked technology. Mark recognition technology is more simple and more mature, but is limited to a particular need to use the mark. No mark recognition technology uses a natural thing to replace special mark, application of the more flexible, but I want to be able to correctly capture the suspicious object useful feature points. Augmented Reality existing library of development or development platform, the main purpose of providing augmented reality application development, do not provide amplification verify unmarked reality technology Research Methodology. Therefore, this paper discusses how to integrate in-depth study and other ARToolKit, OpenCV and OpenGL libraries, design unmarked Augmented Reality integration platform. This platform functions and modular design, in order to provide a more fle