Single Depth View Based Real-Time Reconstruction of Hand-Object Interactions

Reconstructing hand-object interactions is a challenging task due to strong occlusions and complex motions. This article proposes a real-time system that uses a single depth stream to simultaneously reconstruct hand poses, object shape, and rigid/non-rigid motions. To achieve this, we first train a joint learning network to segment the hand and object in a depth image, and to predict the 3D keypoints of the hand. With most layers shared by the two tasks, computation cost is saved for the real-time performance. A hybrid dataset is constructed here to train the network with real data (to learn real-world distributions) and synthetic data (to cover variations of objects, motions, and viewpoints). Next, the depth of the two targets and the keypoints are used in a uniform optimization to reconstruct the interacting motions. Benefitting from a novel tangential contact constraint, the system not only solves the remaining ambiguities but also keeps the real-time performance. Experiments show that our system handles different hand and object shapes, various interactive motions, and moving cameras.

A Kinematic Analysis Method of Double Roller Tripod Joints Based on the Principle of Conjugate Surfaces

To systematically investigate the contact constraint relationships and the influences of fit clearances on the kinematic performances of a double roller tripod joint (DRTJ), a method for the kinematic analysis of DRTJs is proposed based on the principle of conjugate surfaces. In the proposed method, the constraint relations between rollers and tracks as well as between rollers and trunnions are firstly derived based on the principle of conjugate surfaces. Then, according to the constructed constraint relationships, the kinematic analysis model of a DRTJ considering the influences of fit clearances is established. Next, the effectiveness of the proposed method is validated by measuring the relative displacements and angles between rollers and tracks via experiment. Finally, kinematic analyses are carried out and the main results show that the relative pitch angle between rollers and tracks are always kept as zeroes under any working condition by designing the shapes of rollers' outer surfaces to be semi-toroid. The fit clearances have little influence on the kinematic performances of the DRTJ, thus proper fit clearances between rollers and tracks as well as between rollers and trunnions can be designed to improve the lubricating conditions of the DRTJ.

Influence of Constraint Method of Elastic Diaphragm on Numerical Simulation of Pressure-Compensating Emitter

HighlightsDeformation of the elastic diaphragm can affect the flow field characteristics of a pressure-compensating emitter.The contact constraint method’s simulation of the elastic diaphragm were consistent with its actual deformation.The contact constraint method can reduce the relative errors between simulated and experimental results.Abstract. Numerical simulation is an important method for revealing the working principle and optimizing the design of drip emitters. The hydraulic performance of pressure-compensating (PC) emitters is determined by the interaction between the elastic diaphragm and the flow field; therefore, correct deformation of the elastic diaphragm is one of the factors determining the accuracy of numerical simulation of PC emitters. This study investigated the effects of three constraint methods of the elastic diaphragm on the numerical simulation of PC emitters. The three methods were fully fixed constraint (FFC), upper surface fixed constraint (UFC), and contact constraint (CC). Fluid-structure interaction (FSI) simulation was used to analyze the deformation characteristics of the elastic diaphragm, the flow field characteristics, and the flow rate of the PC emitter. The simulated diaphragm deformation and flow rates were compared with the results of a visual experiment and a hydraulic performance experiment, respectively. The simulation results showed that the constraint method affected the diaphragm deformation and flow field of the PC emitter. In comparing the simulation results with the experimental results, the CC method had the highest accuracy among the three constraint methods, but an extremely long computation time was required. The FFC method had the lowest accuracy but required less computation time. The accuracy of the UFC method was lower than CC and higher than FFC, but its computation time decreased by 60.03% compared with CC. This study provides a foundation for further research on the numerical simulation and design of PC emitters. Keywords: Constraint method, Deformation characteristics, Flow field characteristics, Fluid-structure interaction, Visual experiment.

A study of contact methods in the application of large deformation dynamics in self-contact beam

This paper introduces a procedure in the field of computational contact mechanics to analyze contact dynamics of beams undergoing large overall motion with large deformations and in self-contact situations. The presented contact procedure consists of a contact search algorithm which is employed with two approaches to impose contact constraint. The contact search task aims to detect the contact events and to identify the contact point candidates that is accomplished using an algorithm based on intersection of the oriented bounding boxes (OBBs). To impose the contact constraint, an approach based on the complementarity problem (CP) is introduced in the context of beam-to-beam contact. The other approach to enforce the contact constraint in this work is the penalty method, which is often used in the finite element and multibody literature. The latter contact force model is compared against the frictionless variant of the complementarity problem approach, linear complementarity problem approach (LCP). In the considered approaches, the absolute nodal coordinate formulation (ANCF) is used as an underlying finite element method for modeling beam-like structures in multibody applications, in particular. The employed penalty method makes use of an internal iteration scheme based on the Newton solver to fulfill the criteria for minimal penetration. Numerical examples in the case of flexible beams demonstrate the applicability of the introduced approach in a situation where a variety of contact types occur. It was found that the employed contact detection method is sufficiently accurate when paired with the studied contact constraint imposition models in simulation of the contact dynamics problems. It is further shown that the optimization-based complementarity problem approach is computationally more economical than the classical penalty method in the case of studied 2D-problems.