scispace - formally typeset
Search or ask a question

Showing papers by "Matthew Turk published in 1999"


Proceedings Article
01 Jan 1999

2,010 citations


Proceedings ArticleDOI
01 Dec 1999
TL;DR: An algorithm for real-time tracking of articulated structures in dense disparity maps derived from stereo image sequences is presented, able to successfully track upper human body motion in real time and in the presence of self-occlusions.
Abstract: In this paper, we present an algorithm for real-time tracking of articulated structures in dense disparity maps derived from stereo image sequences A statistical image formation model that accounts for occlusions plays the central role in our tracking approach This graphical model (a Bayesian network) assumes that the range image of each part of the structure is formed by drawing the depth candidates from a 3-D Gaussian distribution The advantage over the classical mixture of Gaussians is that our model takes into account occlusions by picking the minimum depth (which could be regarded as a probabilistic version of z-buffering) The model also enforces articulation constraints among the parts of the structure The tracking problem is formulated as an inference problem in the image formation model This model can be extended and used for other tasks in addition to the one described in the paper and can also be used for estimating probability distribution functions instead of the ML estimates of the tracked parameters For the purposes of real-time tracking, we used certain approximations in the inference process, which resulted in a real-time two-stage inference algorithm We were able to successfully track upper human body motion in real time and in the presence of self-occlusions

103 citations




01 Jan 1999
TL;DR: An algorithm for real time 3-D tracking of articulated structures in stereo image sequences that can be captured by an inexpensive commercially available system that also computes the dense disparity map in real time is presented.
Abstract: In this paper, we present an algorithm for real time 3-D tracking of articulated structures in stereo image sequences. These sequences can be captured by an inexpensive commercially available system that also computes the dense disparity map in real time. In our algorithm, the tracked object is modeled as a set of articulated 3D blobs, each adhering to a Gaussian distribution. Classification of the disparity map pixels into the segments of the articulated object is based on the maximum likelihood principle with an additional mechanism for filling the missing data created by self-occlusions. The articulation constraints are enforced through an Extended Kalman Filter, which can also be used to model the dynamics of the tracked object.