An approach for estimating 3D body pose from multiple, uncalibrated views is proposed. First, a mapping from image features to 2D body joint locations is computed using a statistical framework that yields a set of several body pose hypotheses. The concept of a "virtual camera" is introduced that makes this mapping invariant to translation, image-plane rotation, and scaling of the input. As a consequence, the calibration matrices (intrinsics) of the virtual cameras can be considered completely known, and their poses are known up to a single angular displacement parameter Given pose hypotheses obtained in the multiple virtual camera views, the recovery of 3D body pose and camera relative orientations is formulated as a stochastic optimization problem. An Expectation-Maximization algorithm is derived that can obtain the locally most likely (self-consistent) combination of body pose hypotheses. Performance of the approach is evaluated with synthetic sequences as well as real video sequences of human motion.
Estimating 3D body pose using uncalibrated cameras
2001-01-01
796471 byte
Conference paper
Electronic Resource
English
Estimating 3D Body Pose Using Uncalibrated Cameras
British Library Conference Proceedings | 2001
|Algorithm for Estimating Mean Traffic Speed with Uncalibrated Cameras
Transportation Research Record | 2000
|Algorithm for Estimating Mean Traffic Speed with Uncalibrated Cameras
Online Contents | 2000
|Estimating Anthropometry and Pose from a Single Uncalibrated Image
British Library Online Contents | 2001
|Flexible 3D models from uncalibrated cameras
British Library Online Contents | 1996
|