本文目的是利用機械手臂搭配智慧視覺系統實現路徑規劃及夾取功能。主要方式是利用影像資訊針對洞口的形狀以及被夾取物體的形狀，計算應旋轉的角度。再將目標影像的質心位置轉換為手臂的座標系統，並設定夾取順序，同時進行路徑規劃與建立數據庫。機器手臂則以數據庫所規劃之路徑進行作動，進而完成夾取有色積木投入相同顏色與形狀洞口之功能。實驗結果證實本文所採用的影像辨識技術可以完整呈現機器手臂路徑規劃之功能。

In this paper, we make use of an intelligent vision technique to implement the robot path planning and the action of grasping objects. The image-gap information between the captured body and the hole is applied to the calculation of rotation angle. Moreover, the centroid point belongs to the identified geometry in the image is transformed into the coordinate system of the robotic arm such that the grasping sequence is set. Simultaneously, we create the robot path. Furthermore, by analysis of the database, the robotic arm can grasp objects into the correct hole. The experimental results indicate the practicality of the intelligent vision for robot path planning.