本研究目的在設計一輛三輪式可沿著模擬道路行進之自走車，其任務為追蹤擁有特定顏色，行進於平面模擬道路之動態目標物。本系統可分為三個主要部分。第一部分為遠端監控模組，其包括主控電腦，自走車控制器以及接收自走車端資訊的傳輸設備，其傳輸設備以無線傳輸之方式接收影像，並傳送追蹤指令給自走車。第二部分為影像傳輸模組，其包括於自走車上的CCD攝影機以及主控電腦端的影像處理軟體，透過影像處理軟體對即時影像運算，可求得目標物於影像中的顏色，形狀以及形心位置，利用處理過的影像資訊，在此可做為控制器之輸入。第三部分為自走車模組，利用此模組的微處理器晶片，其主要執行動作包括:根據紅外線感測器之輸出訊號下達循軌指令，轉動車上的平台，帶動鏡頭，鎖定目標以及調變左右驅動輪之轉速，改變行進方向。本研究之控制器為模糊理論和決策樹之整合，並以影像資訊做為決策樹之節點判斷，以不同的情況下，傳達追蹤指令於自走車。在本論文之後會以三個基本案例以及一個特殊案例來證明追蹤車可成功達成任務。

The objective of the research is to design a three-wheeled robot car, which can stay on the guide rail and track a moving target with particular color on the horizontal road-block surface. Three major parts are included in designed system. Part 1 is remote monitoring unit, consisting of a personal computer (PC), robot car command controller, and data transmission devices linking to the mobile robot, which receives target image through radio signals from and sends out tracking commands to the robot car. Part 2 is the image processing unit, consisting of CCD camera on the robot car and a set of image processing software located in the PC, which takes in target images and evaluate target color, shape and center as input to the command controller. Part 3 is the mobile robot unit which has its own on-board processor and acts according to the tracking commands using its infrared sensors to maintain on the guide rail, aiming CCD camera toward the target by rotating the camera holding platform on the robot car, and closing into the moving target by varying the speed of its two rear driving wheels. The tracking controller in the PC is formulated with fuzzy logic integrated with decision tree takes in target image info and produces tracking and following commands to the robot car. 3 basic tracking modes representing most of the situations and one general situation are included and the results of the experiments indicated that the objective of the study can successfully achieve.

【1】汪在洋 “自走車在智慧型空間下之動態目標物追蹤”，大同大學機械工程研究所碩士論文。
【2】郭桓甫”應用及時影像辨識於物體追蹤之無線遙控自走車的設計製作”，國立成功大學航空太空工程學系碩士論文，2006。
【3】陳俊嘉 “即時影像追蹤之自走車設計”，國立中央大學電機工程研究所碩士論文，2007。
【4】A.Dick and M. Brooks. A stochastic approach to tracking objects across multiple cameras. Proceedings of the Australian Conference on Artificial Intelligence, 2004.
【5】蔡嘉宏，”應用影像切割技巧建立室內通道的導航系統”，國立交通大學電機與控制工程研究所碩士論文，2000。
【6】繆紹綱譯，普林斯頓國際有限公司，”數位影像處理”，2004。
【7】井上誠喜，C語言數位影像處理。2000。
【8】郭桓甫，應用即時影像辨識技術於物體追蹤之無線遙控自走車的設計製作。
【9】Zadeh,L.A., “Information and Control,”Fuzzy Set,8,338-353(1965)。
【10】蕭方智，”應用階層粒子法方法於模糊決策樹之研究”元智大學工業與工程管理研究所。
【11】Simon X. Yang，Hao Li and Max Meng. “Fuzzy Control of a behavior- based mobile robot” the 12th IEEE international conference on fuzzy systems, vol.1,25-28 May 2003,pp.319-324
【12】P.Rusu,E.m.Petriu, T.E.Whalen, and A. Connell, and H.J.W. Spoelder, ”Behavior-based neuro-fuzzy controller for mobile robot navigation,” IEEE transactions on instrumentation and measurement, vol.52, Issue 4,Aug.2003, pp,1335-1340
【13】廖名揚，”應用即時影像集模糊PID控制自走車於無線攔截任 務”，國立成功大學航空太空工程學系碩士論文，2007。
【14】簡卓建 應用模糊控制於自走車路徑導引避障之整合設計
【15】張世傑 應用車型機器人之模糊停車控制的設計與實現