From gesture recognition to functional motion analysis: quantitative techniques for the application and evaluation of human motion

The quantification and analysis of human motion is a central focus of many studies in biology, anthropology, biomechanics, human factors and ergonomics. These works are primarily concerned with describing the relationships between structure and function from a quantitative perspective. Recent work has seen the application of functional analysis techniques and their associated models to such diverse areas as surveillance, human computer interaction, and game development by researchers in the areas of computer graphics, robotics, computer vision, and machine learning. The work in this dissertation is motivated by the study and quantification of gesture and human motion. This research explores the characteristics of gesture-based interactions, the development of a gesture recognition tool for virtual reality environments, the quantification and analysis of a sequence of postures as a complete motion, and the application of the motion analysis methods to a lifting and fatigue study. The result is a look at gesture and motion analysis from its role in interaction to its ability to quantify relationships between structure and function.