Real-time motion tracking for dance visualization using Kalman filters

Abstract

This paper explores an application of Kalman filters for real-time dance visualization within the Unity3d framework. Experiments performed with camera-based motion sensors indicate that Kalman filter implementation can benefit the process of content development for real-time visualization of dance performances. Moreover, we showed that Kalman filters could be used for generating real-time visuals of dance performances. While the scope of this project is limited to dance visualizations our prototype can be extended to become a useful tool for practitioners to identify any movement dysfunctions, and to educate patients with musculoskeletal impairments on their movements.