Dongsheng Zhou

An Efficient Method of Key-Frame Extraction Based on a Cluster Algorithm

Abstract This paper proposes a novel method of key-frame extraction for use with motion capture data. This method is based on an unsupervised cluster algorithm. First, the motion sequence is clustered into two classes by the similarity distance of the adjacent frames so that the thresholds needed in the next step can be determined adaptively. Second, a dynamic cluster algorithm called ISODATA is used to cluster all the frames and the frames nearest to the center of each class are automatically extracted as key-frames of the sequence. Unlike many other clustering techniques, the present improved cluster algorithm can automatically address different motion types without any need for specified parameters from users. The proposed method is capable of summarizing motion capture data reliably and efficiently. The present work also provides a meaningful comparison between the results of the proposed key-frame extraction technique and other previous methods. These results are evaluated in terms of metrics that measure reconstructed motion and the mean absolute error value, which are derived from the reconstructed data and the original data.

Motion Key-frames extraction based on amplitude of distance characteristic curve

AbstractThe key frames extraction technique extracts key postures to describe the original motion sequence, which has been widely used in motion compression, motion retrieval, motion edition and so on. In this paper, we propose a method based on the amplitude of curve to find key frames in a motion captured sequence. First we select a group of joint distance features to represent the motion and adopt the Principal Component Analysis (PCA) method to obtain the one dimension principal component as a features curve which will be used. Then we gain the initial key-frames by extracting the local optimum points in the curve. At last, we get the final key frames by inserting frames based on the amplitude of the curve and merging key frames too close. A number of experimental examples demonstrate that our method is practicable and efficient not only in the visual performance but also in the aspect of the compression ratio and error rate.

On asymptotic stability of discrete-time non-autonomous delayed Hopfield neural networks

In this paper, we obtain some sufficient conditions for determining the asymptotic stability of discrete-time non-autonomous delayed Hopfield neural networks by utilizing the Lyapunov functional method. An example is given to show the validity of the results.

Practical analytical inverse kinematic approach for 7-DOF space manipulators with joint and attitude limits

In this study, we propose a practical approach for calculating the analytical inverse kinematic solution for a seven-degrees of freedom (7-DOF) space manipulator with joint and attitude limits. Instead of utilizing traditional velocity-based approaches that limit the ranges of joints by calculating the velocity-level Jacobian matrix, we propose a position-based approach for evaluating the ranges of feasible inverse kinematic solutions. We then search for the optimal solution, which is estimated based on the disturbance that acts on the base of the manipulator to obtain the final solution. First, the concept of the redundancy of manipulators is defined and each joint is parameterized by the redundancy. Second, how the joint limits affect this redundancy is discussed. Third, a practical approach (include the objective function that the author needs to minimize) is proposed for dealing with the inverse kinematic problem of 7-DOF manipulators. Finally, the validity of this approach is verified by numerical simulation.

Forward non-rigid motion tracking for facial MoCap

For the existing motion capture (MoCap) data processing methods, manual interventions are always inevitable, most of which are derived from the data tracking process. This paper addresses the problem of tracking non-rigid 3D facial motions from sequences of raw MoCap data in the presence of noise, outliers and long time missing. We present a novel dynamic spatiotemporal framework to automatically solve the problem. First, based on a 3D facial topological structure, a sophisticated non-rigid motion interpreter (SNRMI) is put forward; together with a dynamic searching scheme, it cannot only track the non-missing data to the maximum extent but recover missing data (it can accurately recover more than five adjacent markers under long time (about 5 seconds) missing) accurately. To rule out wrong tracks of the markers labeled in open structures (such as mouth, eyes), a semantic-based heuristic checking method was raised. Second, since the existing methods have not taken the noise propagation problem into account, a forward processing framework is presented to solve the problem. Another contribution is the proposed method could track facial non-rigid motions automatically and forward, and is believed to greatly reduce even eliminate the requirements of human interventions during the facial MoCap data processing. Experimental results proved the effectiveness, robustness and accuracy of our system.

Interactive traffic simulation model with learned local parameters

In this paper, we present a parameter learning method to reflect the rapidly changing behaviors in the traffic flow simulation process, in which we insert virtual vehicles into the real trajectory data. We come up with a real-virtual interaction model and then we use genetic algorithm to learn some parameters in the model with the purpose to get some specific driving characteristics. Then we propose a real-virtual interaction system to vividly simulate the various interaction behaviors between the real vehicles and the virtual ones. Our results are compared to the existing methods to prove the effectiveness of our presented method.

A Segmented Artificial Bee Colony Algorithm Based on Synchronous Learning Factors

In this paper, we propose a segmented ABC algorithm based on synchronous learning factors (SABC). For the problem of inferior local search ability and low convergence precision in the artificial bee colony (ABC) algorithm, we use the method of synchronous change learning factors for local search. Then under the guidance of the segmented thought, it updates the quality honey greedily. It improves the efficiency of nectar source updating, enhances the local search ability of artificial bee colony. The six standard test functions are chosen to do the simulation experiments. Compared with the other three experiments, the results show that SABC has a significant improvement in the convergence speed and searching optimal value.

Optimal Path Planning for Minimizing Base Disturbance of Space Robot

The path planning of free-floating space robot in space on-orbit service has been paid more and more attention. The problem is more complicated because of the interaction between the space robot and base. Therefore, it is necessary to minimize the base position and attitude disturbance to improve the path planning of free-floating space robot, reducing the fuel consumption for the position and attitude maintenance. In this paper, a reasonable path planning method to solve the problem is presented, which is feasible and relatively simple. First, the kinematic model of 6 degrees of freedom free-floating space robot is established. And then the joint angles are parameterized using the 7th order polynomial sine functions. The fitness function is defined according to the position and attitude of minimizing base disturbance and constraints of space robot. Furthermore, an improved chaotic particle swarm optimization (ICPSO) is presented. The proposed algorithm is compared with the standard PSO and CPSO algorithm in the literature by the experimental simulation. The simulation results demonstrate that the proposed algorithm is more effective than the two other approaches, such as easy to find the optimal solution, and this method could provide a satisfactory path for the free-floating space robot.

A new triangulation algorithm from 3D unorganized dense point cloud

This paper presents an algorithm for triangular mesh generation from unorganized points based on 3D Delaunay tetrahedralization and mesh-growing method. This algorithm requires the point density to meet the well-sampled condition in smooth regions and dense sampling in sections of a great curvature and two close opposite surfaces. The principle of the algorithm is as follows. It begins with 3D Delaunay tetrahedralization of all sampling points. Then extract part of triangles belonging to the surface as the seed facets according to the rough separation characteristics which based on the angle formed by the circumscribing balls of incident tetrahedrons. Finally, the algorithm grows the seed facets from front triangles to all triangles of the surface. This paper shows several experimental results which explain this approach is general and applicable to various object topologies.

Multiswarm particle swarm optimization with transfer of the best particle

We propose an improved algorithm, for a multiswarm particle swarm optimization with transfer of the best particle called BMPSO. In the proposed algorithm, we introduce parasitism into the standard particle swarm algorithm (PSO) in order to balance exploration and exploitation, as well as enhancing the capacity for global search to solve nonlinear optimization problems. First, the best particle guides other particles to prevent them from being trapped by local optima. We provide a detailed description of BMPSO. We also present a diversity analysis of the proposed BMPSO, which is explained based on the Sphere function. Finally, we tested the performance of the proposed algorithm with six standard test functions and an engineering problem. Compared with some other algorithms, the results showed that the proposed BMPSO performed better when applied to the test functions and the engineering problem. Furthermore, the proposed BMPSO can be applied to other nonlinear optimization problems.