Interactive mobile pet game

Abstract

As information technologies have been highly developed, mobile games are also becoming popular and splendid. However, the interface between human and computer has hardly changed and users can not operate mobile games exactly what they need due to the small buttons and small screen. In this paper, we present a new intuitive and interactive interface for mobile games using computer vision techniques optimized for low-performance mobile devices. It can track hand locations to interact with a cyber pet dog and execute several menus to give a command to mobile pet game. We use a HSI skin-color model and a CAMShift algorithm, which are adequate for the performance of mobile devices, in order to find the accurate hand location. And we use the Behavior-based AI method drawn from ethology to generate several actions of the pet dog.