首页--工业技术论文--自动化技术、计算机技术论文--计算技术、计算机技术论文--计算机的应用论文--信息处理(信息加工)论文--模式识别与装置论文

Model-Based Human Motion Analysis

摘要第4-5页
Abstract第5页
1 Introduction第12-22页
    1.1 Background and Significance第12-16页
    1.2 Literature Review第16-18页
    1.3 The Challenges第18-20页
    1.4 Content and Structure of the Study第20-22页
2 Model Matching and Forces Interaction on Ellipsoidal HumanLimb第22-60页
    2.1 Overview第22页
    2.2 Modeling Approach第22-32页
        2.2.1 Ellipsoids第23页
        2.2.2 Simplified Limb Model第23-24页
        2.2.3 Skeleton Modeling第24-26页
        2.2.4 Ellipsoid Modeling第26-27页
        2.2.5 Deformation of Ellipsoid第27-32页
    2.3 Muscle Modeling第32-33页
    2.4 Working and Method第33-34页
    2.5 Description of Model and Derived Equations第34-40页
        2.5.1 Force- Length Relationship F_L~M(L~M)第35-36页
        2.5.2 Force- Velocity Relationship F_V~M(L~M)第36-38页
        2.5.3 Parallel Elastic Element F~(PE)(L_O~M)第38-39页
        2.5.4 Length and Moment Arm of Human Arm Muscles第39-40页
    2.6 Representing Musculotendon Parameters第40-45页
    2.7 Unknown Parameter Estimation Method第45-46页
    2.8 Model Matching第46-51页
        2.8.1 Skin Segmentation第46-47页
        2.8.2 Skeleton Fitting第47-48页
        2.8.3 Matching Approach第48-51页
    2.9 Template Model Matching (Ellipsoidal Fitting)第51-52页
    2.10 Experiment and Findings第52-58页
    2.11 Conclusion第58-60页
3 Tensor Product Human Limb第60-89页
    3.1 From Curves to Surface第60-62页
    3.2 Surface Patch Continuity第62-66页
        3.2.1 Insight to Patch Continuity第62-64页
        3.2.2 The G1 Continuity第64-66页
    3.3 The Model Development第66-72页
        3.3.1 Data Points and Model Association第67页
        3.3.2 The Center of Mass and Geometric Center第67-68页
        3.3.3 Input Three Dimensional Data第68-69页
        3.3.4 Least Mean Square Solution第69-70页
        3.3.5 Measuring Errors第70页
        3.3.6 Fitting A Tensor Product Surface第70-72页
        3.3.7 Surface Generating Algorithm第72页
    3.4 Single Parametric Patch and its Control Points第72-76页
    3.5 Surface Deformation第76-85页
        3.5.1 Simplified Boundary Curve第76-77页
        3.5.2 Maintaining G1第77-78页
        3.5.3 Reverse deCasteljau第78-80页
        3.5.4 Affected Region第80页
        3.5.5 Simplified Boundary Curve Control Point Perturbation第80-83页
        3.5.6 Optimization of boundary curve network第83-85页
    3.6 Experiment and Findings第85-87页
    3.7 Conclusion第87-89页
4 Human Motion Analysis: Full Body Segmentation, PoseEstimation and Parsing第89-125页
    4.1 Overview第89页
    4.2 Feature Extraction and GMM Modeling第89-93页
        4.2.1 Silhouette Chamfer第89-91页
        4.2.2 GMM Based Pixel Intensity Modeling第91页
        4.2.3 GMM Learning and Updating第91-93页
    4.3 Top-Down Inference for Image Segmentation第93-101页
        4.3.1 Markov Random Fields第93-94页
        4.3.2 Treating Image Segmentation as MAP-MRF Inference第94-96页
        4.3.3 Pose-Specific MRF and Non-Rigid Model第96-98页
        4.3.4 Augmenting Pose-Specific Shape Prior to MRF第98-99页
        4.3.5 Graph Cuts for MAP-MRF Inference第99页
        4.3.6 Formulating the Pose Inference Problem第99-100页
        4.3.7 Minimizing Energies Using Dynamic Graph cuts第100-101页
    4.4 Bottom-Up Cues for Body Parsing and Pose Fixing第101-112页
        4.4.1 Representing Natural Sub-Segments第101-103页
        4.4.2 Shape Decomposition Computation第103-105页
        4.4.3 Transform Invariant Body Model第105-107页
        4.4.4 Probabilistic Model for Shape Variance第107-108页
        4.4.5 Body Part Labeling Using Bayesian Similarity第108页
        4.4.6 Problem Formulation第108-109页
        4.4.7 Calculation of Goodness Function第109-111页
        4.4.8 Selection of Optimal Hypothesis第111-112页
    4.5 Recursive Context Reasoning第112-118页
        4.5.1 Updating Shapes and Locations of Body Parts第113-114页
        4.5.2 Parameter Update of the Identified Parts第114-116页
        4.5.3 Prediction of Missing Part's Parameters第116-118页
    4.6 Experiments and Results第118-124页
        4.6.1 Findings from Datasets第118-123页
        4.6.2 Comparison with other Methods第123-124页
    4.7 Conclusion第124-125页
5 Summary, Findings and Recommendation第125-128页
    5.1 Summary and Findings第125-126页
    5.2 Recommendation for Further Research Work第126-128页
Acknowledgements第128-129页
References第129-137页
Appendix第137页

论文共137页,点击 下载论文
上一篇:基于Nutch的分布式主题爬虫研究和实现
下一篇:面向物联网的Modbus协议栈设计与应用