Hand gesticulation interpretation via smart sensing

This paper describes a part of a vision system that interprets hand gesticulation. The system consists of three sub-systems each of which comprises a number of modules targeted at performing specific functions. The modules contain adaptive image processing and pattern recognition algorithms. The first sub-system is a hand-motion tracker that determines hand trajectory in a 3-D space using a sequence of low-resolution images. The second sub-system interprets the hand articulation using a smart sensor that provides selective views of variable resolution. The third sub-system fuses the outputs of the motion tracker and smart sensor and gives the meaning to hand gesticulation. The paper discusses the technical details of the algorithms used for interpreting hand articulation