本文提出一個新的基於開迴路單眼視覺運動控制的即時目標追蹤法。此方法利用粒子濾波技術預測移動目標在影像中之位置；由於粒子濾波的特性，此方法能有效地掌握線性與非線性之運動行為。另外，此方法使用簡單的數學運算將移動目標之影像資訊轉換為其真實座標資訊，所以此方法消耗少量的運算資源。再者，此方法採用單眼視覺法，亦即使用單一攝影機，可以使用較少的硬體資源來實現。此方法會先針對圖像內目標物下一時刻的位置及大小進行估計，再由預測所得之資訊對應至物體真實位置，並控制移動機器人使目標物保持在攝影機之視野中央。本文進行了L 型與S 型追蹤測試，並與卡爾曼濾波法進行比較。由實驗結果顯示，本文所提之方法在L 型實驗中達到良好的追蹤效果，並且於兩種實驗中均優於卡爾曼濾波法。

In this paper, a new method, called the open-loop mono-vision based motion control (OMMC), for real-time target tracking with mobile robots is proposed. In the OMMC, a moving target’s position in an image is predicted by the particle filter technique. Due to the stochastic properties of particle filtering, the OMMC can effectively and accurately cover and handle both linear and nonlinear dynamic motion behaviors. In addition, it uses simple polynomial calculations to transfer a target’s position in an image to its real coordinates, and therefore it requires few software resources for computation purposes. Moreover, the OMMC adopts the monocular vision approach, i.e., it uses a single camera, and therefore it needs few hardware resources for system implementation. Overall, the OMMC, executed by a mobile robot,predicts a moving target’s next position in an image, and calculates its real coordinates.Then, the mobile robot is commanded to move towards the predicted position, so as to keep the target at the camera s central view. Experimental results show that the OMMC well performs in the L-shape tracking experiment, and performs better than the Kalman filter technique whether in the L-shape or S-shape tracking experiment in this study.