以單一攝影機完成同步定位、地圖建置與物體追蹤之非延遲初始化演算法

Abstract

已有不少基於卡爾曼濾波器的研究結果展示了使用單一相機來進行同步定位、建立地圖(SLAM)的可行性。然而，較少研究探討SLAM 在動態環境中的可行性。為了能在動態環境中同時建立靜態與動態地圖，我們提出一個基於卡爾曼濾波器的演算法架構及新的參數表示法來整合移動物體。藉由新的參數表示法，我們的演算法能同時估測環境中的靜態物體及動態物體，而達到廣泛物體的地圖建製(SLAM with generalized objects)。這樣的參數表示法繼承了倒數深度表示法(Inverse depth parametrization)的優點，像是較大範圍的距離估測、較佳的線性化參數表示。目前關於SLAM 在動態環境中的研究，皆需要數筆測量以確保物體的靜止性質，再延遲的進行物體初始化。而我們的參數表示法允許無延遲的物體初始化，使得我們的演算法能利用每一筆的測量而獲得更好的估測。同時，我們也提出了一個低運算量的動態、靜態物體分類演算法。模擬實驗顯示了我們演算法的準確性。而真實環境實驗也顯示了我們的演算法能在室內動態環境成功的進行廣泛物體的地圖建製(SLAM with generalized objects)。

RECENT works have shown the feasibility of the extended Kalman filtering(EKF) approach on simultaneous localization and mapping (SLAM) with a single camera. However, few approaches have addressed the solutions for the insufficient of SLAM to deal with dynamic environments. For accomplishing SLAM in dynamic environments, we proposed a unified framework based on a new parametrization for both static and non-static point features. By applying the new parametrization, the algorithmis able to integratemoving features and thus achieve monocular SLAM with generalized objects. The new parametrization inherits good properties of the inverse depth parametrization such as the ability to adopt large range of depths and better linearity. In addition, the new parametrization allows undelayed feature initialization. Contrary to the existing SLAM algorithms with delayed initialization approach which takes some measurements for the classification usage, our SLAM with generalized objects algorithmwith undelayed initialization would utilize each measurement on point features for filtering and has a better estimation of the environment. A low computational classification algorithmto distinguish static andmoving features is also presented. Simulations shows high accuracy of our classification algorithm and estimation about features. We also demonstrate the success of our algorithm with real image sequence captured from an indoor environment.