視覺在人的生活中扮演很重要的角色，本論文旨在利用影像改善輪型機器人 ( KNR ) 之目標追蹤能力，為達成此目標，本研究將模糊理論植入影像處理中，利用模糊集合、建立模糊規則庫、模糊推理及解模糊化，設計一個簡易模糊控制器 ( Fuzzy Controller, FC ) ，此控制器之程式則在LabVIEW環境下發展，計算左右輪馬達之轉動速度，降低追蹤誤差。此研究使用LabVIEW 的Vision Acquisition及Vision Assistant做影像前置處理，因LabVIEW的精準與即時特性，已完成研究目標。本研究包括以下步驟，分述如下：
(1) 平台建構。
(2) 模糊控制器設計。
(3) 撰寫LabVIEW 影像處理程式：運用LabVIEW程式上的Vision Acquisition啟動開啟機器人 ( KNR ) 上的HD攝影機，確定影像的像素及是否捕獲到影像，進一步運用Vision Assistant對影像進行前置作業（二值化）和後置作業（圖像辨識）以便取得座標值。
(4) 撰寫模糊控制器程式：透過圖像辨識後之座標值，計算左右輪馬達之轉動速度，並以影像保持固定距離以防止物體與機器人發生衝撞，使輪型機器人具有影像追蹤的能力。
(5) 程式除錯與功能測試。
本研究利用LabVIEW與模糊演算方法，已完成了影像辨識與目標追蹤，並已成功地運用在輪型機器人之目標追蹤上。
Vision plays an important role in daily life. In this thesis, the vision is sued to improve the target tracking ability of the wheeled robot (KNR). To achieve this goal, the fuzzy theory is embedded into image processing in the study. The fuzzy set, fuzzy rule base, fuzzy inference and defuzzification are utilized to design a fuzzy controller (FC). The program of the proposed controller is developed under the LabVIEW environment to calculate the speeds of right and left wheels, respectively. The Vision Acquisition and Vision Assistant packages are used to pre-process the captured image. The developing stage contains following five steps(1) Establish the wheeled robot (KNR).
(2) Design the fuzzy controller.
(3) Develop the image processing program under LabVIEW environment: The functions include (a) Triggering the camera of the controlled wheeled robots (KNR); (b) Capturing the image on fixed interval time; (c) Pre-processing (binary) and post-processing (pattern recognition) to obtain the axis values.
(4) Write a fuzzy controller program: Through the coordinate values of image recognition, calculate the rotational speed of the left and right wheel motor, and to maintain a fixed distance from the image in order to prevent the occurrence of a collision object and robots, wheeled robot with the ability to make video tracking.
(5) Program debug and test.
In this thesis, the technologies of fuzzy theory and image processing have been combined successfully to recognize the pattern and drive the wheels of the robot. And also the LabVIEW program has been successfully executed to capture the image, access the axis values and convert driving voltages of the wheeled robot (KNR).
