Functional body mesh representation - a simplified kinematic model

This paper describes Functional Body Mesh a simplified representation of a body kinematic structure intended to form a framework for marker-wise signal processing of recorded motion trajectories of motion capture of virtually any vertebrates. The parent-child and sibling relationships are inferred on a coherence of movement and constancy of distances. There is proposed a GMM based method for inferring FBM segments from the raw motion capture data based on a relationship between the points. For creating groups representing specific body parts we propose an incremental multicriterial clustering algorithm employing Gaussian mixture models. To infer body parts hierarchy we propose utilizing a consensus method.