視覺與力覺整合於機器人遠端操作之應用

Abstract

由於網路與微處理器的快速進步，使得遠端操作的性能大幅提高，為了更進一步提升遠端操作的效能，讓使用者的感受能更加真實，越來越多的研究投入力覺於遠端操作中的應用，例如力覺與觸覺的回授、虛擬導引、虛擬支架(virtual fixture)等。在本論文中，我們實現了一個遠端呈現與操作系統，結合了視覺與力覺，並能隨著使用者的操作而改變引導方式的虛擬導引(virtual guidance)系統；我們藉由遠端架設的攝影機對欲操作的物件定位，透過力感測器對物件估測其物理參數，然後將這些資訊傳入近端的虛擬實境引擎以及虛擬力回饋演算法中，產生一同時利用視覺和力覺的虛擬導引來協助使用者，讓使用者能夠自然、有效地進行操作以達成目標。

Rapid development in network and microprocessor highly enhance the capability of teleoperation systems. And, dramatic increase in haptic devices greatly raises the efficiency in teleoperation and the fidelity on reality. In this thesis, we integrate both vision and force together for the networked VR-based telerobotic system developed in our laboratory. The main purpose of the integration is to provide virtual guidance, which may assist the manipulation of the operator. This virtual guidance, different from previous approaches, can adjust to the location and physical property of the object, as the status of the manipulation. Experiments are performed to demonstrate the effectiveness of the proposed system.