Design of embedded gesture recognition system for robotic applications

Recent developments of system-on-chip (SOC) technologies have found a lot of applications in robotic embedded system design. Powerful miniature SOC boards allow development of mobile robots with advanced onboard computation performance executing complex data processing and control algorithms. This paper presents a preliminary results of an embedded system design of a mobile robot platform implementing gesture recognition and control. The system is designed using a BeagleBoard-xM SOC board and an Asus Xtion Live 3D camera. System implementation and analysis of a human hand detection and tracking from depth images are discussed.