Dynamical System Representation, Generation, and Recognition of Basic Oscillatory Motion Gestures - Automatic Face and Gesture Recognition, 1996., Proceedings of the Second International Conference

We present a system for generation and recognition of oscillatory gestures. Inspired by gestures used in two representative human-tohuman control areas, we consider a set of oscillatory motions and refine from them a 24 gesture lexicon. Each gesture is modeled as a dynamical system with added geometric constraints to allow for real time gesture recognition using a small amount of processing time and memory. The gestures are used to control a pan-tilt camera neck. We propose extensions for use in areas such as mobile robot control and telerobotics. 1 Developing A Gesture Lexicon. Sociological and biological research on human created gestures suggests that while gestures have standard meanings within a society, no known body motion or gesture has the same meaning in all societies [l]. Even in American Sign Language, few signs are so clearly transparent that a non-signer can guess their meaning without additional clues [6]. Thus we are free to create gestures for device control as we see fit. 1.1 Examples of a Human Gestural Control Environment. Two areas in which gesture languages have developed to communicate commands are crane and excavator control’ and runway traffic control. A sample set of crane control gestures, shown in figure 1 [3], is composed of oscillating planar motions, that is, circles or back-and-forth lines made in two dimensions in real world three dimensional space. Certain gestures used to signal aircraft on a runway are also planar oscillators [IO]. The use of gestures in these environments shows that oscillatory motions are useful for several reasons. First, oscillatory motions are recognizable by other humans and used in critical and potentially dangerous areas. Second, humans can easily and consistently make oscillatory motions. Third, some oscillatory gestures ‘Thanks to Prof. Louis Whitcomb for suggesting this application. Turn Turn Right Figure 1: Sample crane control gestures. have time dependent content which, can be created and understood by humans. For example, the “travel ahead” gesture’s circular velocity is increased when a faster response is desired. 1.2 An “Oscillating Motion” The oscillating circles and lines used in the crane and runway lexicons form the basis of the gestures used in our system. When the geometric features of size and direction are added, the lexicon is expanded to encompass 24 gestures (figure Gesture Lexicon.