本論文設計一台兩輪平衡車，並且設計了多種平衡控制器方法，在控制架構中又可分為並聯式控制架構及串聯式控制架構，共設計五種方法，分別為並聯傳統PD型、並聯智慧型、並聯自我學習型、串聯傳統PID型以及串聯智慧型控制方法，再透過一些測試場景驗證不同控制器所控制之兩輪平衡車的穩定性與強健性。接著，為了使兩輪平衡車能夠更全方位的朝向居家服務機器人的目標邁進，本論文為兩輪平衡車加裝無線攝影機，利用MATLAB影像處理工具箱進行影像處理工作，並透過藍芽模組將影像資訊傳送制Arduino mega 2560開發板中，使得兩輪平衡車能夠執行循跡控制之任務。另外，為了讓兩輪平衡車能夠避開障礙物，因此加裝了紅外線感測器，賦予所設計之兩輪平衡車擁有更廣泛之應用。 The subject of the thesis is to present the new designing for a two-wheeled self-balancing robot with image detection system. The robot can balance by itself with only two wheels, this thesis proposes two controller structures, one is parallel structure the other one is cascade structure. In the thesis, a slope is applied to prove the stability and robustness of two-wheeled self-balancing robot. Nowadays, the home service robot develop quickly and has more functions, therefor the camera was installed on the top of robot, and use the image processing toolbox of MATLAB to trace a black line, and design a fuzzy steering controller to complete the tracing mission. To implement the avoid obstacles mission. In the thesis also installed infrared sensor, thus the two-wheeled self-balancing robot can avoid obstacles. Finally, the experimental results shows that the two-wheeled self-balancing robot not only can authenticate both in the quality of the balance and steering controller systems but also present the application area for the new function of robot.