Human body modelling for people localization and tracking from real image sequences
Alessandra Teseï · 1995
A method for tracking human motion in 3D real scenes by means of Kalman filtering is presented; it is based on the introduction of a suitable mathematical model for human body moving in a 3D scene and interacting with the surrounding environment. With respect to previous study, the developed refined static and dynamic models allow the system to get accuracy and robustness even in non-controlled real-life working conditions. The method can be used in surveillance systems devoted to assure people safety in public areas.