This paper describes a n approach t o the extraction of articulated objects which will be used f o r gait analysis. In most medical applications markers are used t o determine trajectories of different body parts. This approach works without any markers. Monotony operators which compute the displacement vector field are used t o initialize a contour based tracking algorithmcalled active rays -f o r several body parts which are important f o r gait analysis. T h e contours of differe n t parts of the h u m a n body are extracted and tracked. These parts are approached by simple 3 0 geometric objects (blocks), which 3 0 position and m o t i o n are estimated f o r the each image of the image sequence. Then, the trajectories of the moving parts represented b y the 3D blocks can be determined and used f o r classification of different gait disorders.

In this paper we describe a system for automatic gait analysis. Different kinds of human gait are recognized using sequences of grey-level images. No markers are needed to get the trajectories of different body parts. The tracking of body parts and the classification are based on statistical models. We model several body parts and the background as mixture densities. The positions are determined iteratively, we begin with the most stable part to find. The anatomy of a human body restricts the area to search for the next one. From the trajectories, features for gait analysis are derived. These are used to train hidden Markov models (HMMs), one HMM for each kind of gait. The authours are members of the Graduiertenkolleg 3-D image analysis and synthesis sponsored by the Deutsche Forschungsgemeinschaft (DFG).