本論文以增加狀態卡爾曼過濾器(augmented state Kalman filter, ASKF)建立視覺式同時定位與建圖系統。此SLAM系統的感測器使用單眼視覺系統為唯一的感測裝置，搭配加速強健特徵(speeded-up robust features, SURF)的偵測方法進行環境中影像特徵的偵測，並且依據影像訊息求算特徵在空間中的三維座標，以建立SURF特徵式地圖。對於特徵式地圖中有關資料關聯、資料新增、刪除與更新等問題，皆有提出本研究之解決策略。
本研究對SLAM系統所發展出的相關技術，對於地圖資料庫中的3D影像特徵，透過狄勞尼三角化(Delaunay triangulation)之理論建立有限的馬賽克(mosaic)網格，實現單眼視覺系統於運動中探知結構(structure from motion, SFM)的分析。
最後，本論文也提出實測範例，驗證所提方案的可行性。首先，使用單眼視覺實現具備未知輸入之系統的同時自我定位與特徵式地圖建立，同時也解決單眼視覺的影像深度量測問題。最後，整合ASKF估測方法、特徵式地圖管理策略與馬賽克方法，實現單眼視覺式在3D空間中同時定位、建圖與運動中探知結構之任務。 In this thesis, the visual simultaneous localization and mapping (SLAM) is established by using the augmented state Kalman filter (ASKF). The theory and methodology of vision sensing and state estimation system will be investigated in this thesis. First of all, the visual system is the only sensing device in the system. Meanwhile, the detection of image features, the speeded-up robust features (SURF) with high-dimensional description vectors are utilized to describe the map features, and build the feature-based map. In data association, a tracking window is planned based on the prediction of map features in spatial location, and then the nearest neighbor method is employed to match the high-dimensional descriptor vector of the measured features with that of the features in the map. Secondly, the ASKF is employed to predict and update the states of the robot and the features recursively. Furthermore, the procedures of adding, erasing and updating the data of the SURF in the map are planned. Finally, we use feature-based map to establish mosaicing by using Delaunay triangulation and present the structure from motion of monocular vision.
