本篇論文主要是設計與製作一部智慧型輪型機器人,可以自動辨識並撿拾乒乓球。此機器人利用影像處理技術及馬達控制的技術來實現以上功能功能。首先根據從攝影機所得到的顏色資料讓機器人找尋乒乓球,再計算出乒乓球的中心位置,使得機器人可以往乒乓球方向前進,然後撿拾起該球。當完成任務後機器人還會自動歸位。馬達驅動的模糊控制、搜尋球的機制、避障機制以及回歸放球的機制均再本論文中實現。最後根據機器人的數學模型,設計線性控制方法,再使用MATLAB來進行模擬,以比較馬達驅動線性控制與模糊控制的性能,模擬結果證實模糊控制是比線性控制具有較佳的控制效果。 This thesis designs an intelligent mobile robot that can recognize and collect ping-pong balls automatically. The main techniques on the robot include image processing and motor control. According to the color data of the image getting from the webcam, the robot can find the ping-pong balls. Then, the center location of the ping-pong ball is calculated such that the robot can go forward to the ping-pong ball and pick up it. After the task is completed, the robot will return to the location of ball collection station. The robot moving is based on the motor driving and the motor driving control is based on the fuzzy control techniques to achieve ball searching, obstacle avoidance, and returning to the collection box. Furthermore, the simulation results of the linear controller and fuzzy controller for the robot moving control are compared; it is seen that the fuzzy controller has much better performance than the linear controller.