A framework for motion capture system design using Cramer-Rao lower bound

Motion capture system design is becoming an important subject since the number of its applications is steadily growing and new technologies are introduced into the market. This paper presents a theoretical approach based on Cramer-Rao Lower Bound allowing the designer to choose a configuration (i.e. modality, placement) of sensors, compare different approaches and validate the efficiency of the estimation algorithm to be used in estimating the pose of a subject. An adapted Cramer-Rao bound expression has been derived, and its computational algorithm is presented. The optimization of the design of a human machine interface system based on arm, forearm and hand pose capture using magnetic sensors is then presented as an example.