Capturing and animating occluded cloth

Abstract

Figure 1: We reconstruct a stationary sleeve using thousands of markers to estimate the geometry (texture added with bump mapping). AbstractWe capture the shape of moving cloth using a custom set of color markers printed on the surface of the cloth. The output is a sequence of triangle meshes with static connectivity and with detail at the scale of individual markers in both smooth and folded regions. We compute markers' coordinates in space using correspondence across multiple synchronized video cameras. Correspondence is determined from color information in small neighborhoods and refined using a novel strain pruning process. Final correspondence does not require neighborhood information. We use a novel data driven hole-filling technique to fill occluded regions. Our results include several challenging examples: a wrinkled shirt sleeve, a dancing pair of pants, and a rag tossed onto a cup. Finally, we demonstrate that cloth capture is reusable by animating a pair of pants using human motion capture data.

Deformable 3D Fusion: From Partial Dynamic 3D Observations to Complete 4D Models

“…To this end, we made use of three point-cloud sequences (the pants sequence of [34], and the man and woman sequences of [9]), and of two sequences captured with a Kinect. The three point-cloud sequences allow us to perform quantitative evaluations, while the Kinect sequences illustrate the use of our approach on depth measurements.…”

Estimation and utilization of articulations in recovering non-rigid structure from motion using motion subspaces

“…If one considers the data coming from human body motion, cloth deformation or medical imaging, the motion can be broken down into three components -rigid, articulated and non-rigid. For example, in the cloth data provided by [10], a person wearing a loose fitting cloth is running. The motion of the legs cause articulated motion whereas the waving and ripples in the cloth cause non-rigid deformation.…”

“…Three are presented in this section. The examples correspond to paper deformation [4], a synthetic two-link object and human running motion captured sequence [10]. The paper data contains an image sequence of a paper sharply bending at two locations.…”

Simple data-driven modeling of brushes

“…Data-driven simulation: Data-driven techniques have been used in other areas of simulation, for instance in animating cloth [Cordier and Magnenat-Thalmann 2004;White et al 2007]. Though in many ways more general than the technique we present, these techniques are not particularly suited to the task of simulating brushes in real time.…”