Dynamic Human Pose Estimation using Markov Chain Monte Carlo Approach
WACV-MOTION '05 Proceedings of the IEEE Workshop on Motion and Video Computing (WACV/MOTION'05) - Volume 2 - Volume 02
Recovering 3D Human Pose from Monocular Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
This paper proposes a motion-based search strategy for human pose tracking from a monocular image sequence or video stream. The human pose estimation method compares the image features between 3D human model projections and real human images. The human pose is estimated from the configuration that generates the best match. When searching for the best matching configuration with respect to the input image, the search region is determined from the estimated 2D image motion and then search is performed randomly for the body configuration conducted within that search region. As the 2D image motion is highly constrained, this significantly reduces the dimensionality of the feasible space. This strategy has two advantages. First, the motion estimation leads to an efficient allocation of the search space, and second, the pose estimation method is adaptive to various kinds of motion.