An unsupervised approach to segment the swimmer in hydrodynamic video sequences is presented in this paper. Detecting and segmenting objects in hydrodynamic scenes is challenging as both the desired foreground and unwanted background are driven by complex nonlinear dynamics. These dynamics typically comprise a chaotic mix of vortices, flows and turbulence. Medium introduced reflection and light diffraction creates the camouflage, rendering the swimmer detection even more difficult. On the other hand, spatial regularities and temporal distortions within these sequences often exhibit signature-like patterns. The novelty of our approach is to combine texture modeling and photometric features in order to identify the foreground subject, which lacks the visual characteristics of hydrodynamic content. Segmentation results are then temporally propagated from one frame to another to reduce the effects of camouflage and confine the segmentation to desired region. Simulation results demonstrate ability of the algorithm to separate human body in hydrodynamic sequences for silhouette based underwater motion capture and analysis.

In this paper, we propose a novel two-level regression method for computing body mass distribution from a database of X-ray images, without scanning the new subject. Our approach first selects a suitable sample from the image database by minimizing a distance function based on the relationships between the new subject's body measurements and those of sample subjects. The X-ray image of the new subject is then predicted from the sample image using a feature-based transformation. Body mass distribution is computed directly from the predicted X-ray image. Our results surpass the accuracy of commonly used mass distribution regression methods in biomechanics literatures. In addition, by not scanning the new subject, we avoid all the radiation and cost involved in X-ray absorptiometry.

We address the volumetric reconstruction problem that takes as input a series of orthographic multi-energy x-ray images, producing as output a reconstructed model space consisting of uniform-size mass density voxels. Our approach solves the non-linear constrained optimization for mulation problem by constructing a compliant estimate of volumetric distribution, subject to projective and domain constraints, and minimizes variational irregularities. To resolve the inherent ambiguities of single-view formulation, an optional shape model may be introduced to aid the reconstruction process. We demonstrate our method’s practical usage as a new in-vivo method for estimating three dimensional body segmental compositions, and compare its results with those of existing methods.

A fully automatic 3D-2D articulated registration algorithm for aligning a whole-body human geometry to a 2D X-ray image is presented. Domain prior in the form of a hierarchical model encapsulating the kinematic structure of the human body is exploited. A deformation operator is then defined with respect to the hierarchical model. An iterative search using ICP is performed to estimate the optimal deformation parameters. Empirical results indicate good matching rates of 94% and above. Requiring minimal user intervention, the method is suited for large-scale data acquisition in human anthropometric studies.

Despite leaps in motion capture technology, the dichotomy between unencumbered vision-based motion recovery and the prevailing marker-assisted motion capture solution remains largely pertinent. This paper bridges the gap by introducing an attachment-free, full-body motion capture technique that employs multiple high-speed cameras and a customized bodysuit. The flexible bodysuit allows free movements while providing marker data for guided reconstruction of pose and motion. Our method overcomes several practical problems in existing systems, including long preparation time and displaced markers. The network of markerized pattern provides connectivity information and enhances robustness in feature tracking, model matching and 3D reconstruction.

Efficient numerical techniques developed in the field of computer graphics are able to simulate compellingly realistic simulations of interactions between solids and fluids. These techniques are less used for engineering applications due to errors inherent in the systemic approximations. The errors, manifesting as excessive damping, are expected to change the nature of steady-state fields and nullify the typical engineering static flow analysis. On the other hand, near-body solid/fluids interaction are affected to a lesser degree since errors originating from dissipation are more severe when accumulated over time. Although it is generally not possible to numerically validate unsteady, high-velocity vector fields, our investigations show that near- body solid/fluid dynamics converges with respect to parametric refinements. Though far from a correctness proof, the numerical convergence of near-body quantities suggests the applicability of the method to certain classes of analysis and visualization where near-body characteristics are of greater concern. This article applies the semi-Lagrangian stable fluids method to biomechanical hydrodynamics visualization. The near-body surface dynamics provide meaningful information for rendering visuals that are intuitive to the streamlined flow characteristics surrounding the body. The techniques are applied to the visualization of active and passive resistive forces on the body in a video-based capture of an immersed dolphin kick.

Optimization is a promising way to generate new animations from a minimal amount of input data. Physically based optimization techniques, however, are difficult to scale to complex animated characters, in part because evaluating and differentiating physical quantities becomes prohibitively slow. Traditional approaches often require optimizing or constraining parameters involving joint torques; obtaining first derivatives for these parameters is generally an O ( D 2 ) process, where D is the number of degrees of freedom of the character. In this paper, we describe a set of objective functions and constraints that lead to linear time analytical first derivatives. The surprising finding is that this set includes constraints on physical validity, such as ground contact constraints. Considering only constraints and objective functions that lead to linear time first derivatives results in fast per-iteration computation times and an optimization problem that appears to scale well to more complex characters. We show that qualities such as squash-and-stretch that are expected from physically based optimization result from our approach. Our animation system is particularly useful for synthesizing highly dynamic motions, and we show examples of swinging and leaping motions for characters having from 7 to 22 degrees of freedom.