Performability models for designing disaster tolerant Infrastructure-as-a-Service cloud computing systems

Abstract

Natural and friendly interface is critical for the development of service robot. Gesture-based interface offers an easy and efficient way to interact with robots. In this paper, we implement a Gesture-based interaction system for manipulating service robot. A robust algorithm based on Support Vector Machines (SVM) for recognizing hand gesture is proposed. Our method first detects the hand region by a two-step strategy based on skin color as well as depth images. The Hu moment invariants are then extracted from the hand region. After that, an SVM classifier is trained and is employed to recognize the hand gestures, Finally the interaction system server translates the gesture into pre-defined commands and sends it to the service robot “SIATRob”. The accuracy and effectiveness of the algorithm are examined over several videos with different participants under complex background, changing lights as well as image noises. Experimental results show that our method can reliably and accurately recognize hand gestures in real time.