by Kwan Yuk-choi James, Ng Yuk-kau Sam. / Bibliography: leaves 111-112 / Thesis (M.B.A.)--Chinese University of Hong Kong, 1982

Choi, Yuet Kei. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (leaves 68-72). / Abstracts in English and Chinese. / Acknowledgement --- p.i / Abstract --- p.ii / Contents --- p.vi / List of Figures --- p.viii / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Previous Works --- p.3 / Chapter 1.1.1 --- Template deformation --- p.3 / Chapter 1.1.2 --- Mesh partitioning --- p.3 / Chapter 1.1.3 --- Mesh Smoothing --- p.4 / Chapter 1.2 --- Overview of the approach --- p.5 / Chapter 1.3 --- Thesis outline --- p.7 / Chapter 2. --- Global Deformation --- p.8 / Chapter 2.1 --- The closet point method --- p.10 / Chapter 2.1.1 --- Computational complexity --- p.11 / Chapter 2.2 --- Deformation Techniques --- p.12 / Chapter 2.2.1 --- Existing deformation method --- p.12 / Chapter 2.2.2 --- Radial Basis Functions (RBFs) --- p.14 / Chapter 2.2.2.1 --- Computational complexity --- p.18 / Chapter 2.2.3 --- Result --- p.18 / Chapter 2.3 --- Face flip prevention --- p.20 / Chapter 2.3.1 --- Detection of the flipped face --- p.22 / Chapter 2.3.1.1 --- Common approach: --- p.22 / Chapter 2.3.1.2 --- Our Approach --- p.23 / Chapter 2.3.1.3 --- Comparisons of the face flip detection method: --- p.26 / Chapter 2.3.2 --- Local Subdivision --- p.27 / Chapter 3. --- Partitioning of the mesh --- p.29 / Chapter 3.1 --- Existing method --- p.29 / Chapter 3.2 --- Our approach --- p.31 / Chapter 3.3 --- Computational complexity --- p.34 / Chapter 4. --- Mesh smoothing algorithm --- p.35 / Chapter 4.1 --- The Laplacian flow method --- p.36 / Chapter 4.2 --- The mean-curvature method --- p.40 / Chapter 4.3 --- Our Approach --- p.43 / Chapter 4.3.1 --- The modified mean-curvature method --- p.43 / Chapter 4.3.2 --- The modified Laplacian flow method --- p.45 / Chapter 4.3.3 --- Feature constraints --- p.47 / Chapter 4.3.4 --- Computational complexity --- p.47 / Chapter 4.4 --- Comparison of the mesh smoothing approach --- p.48 / Chapter 5. --- Implementation and Results --- p.51 / Chapter 5.1 --- Construction of the template mesh and boundary curves --- p.51 / Chapter 5.2 --- Selection of the corresponding vertex pairs --- p.52 / Chapter 5.3 --- Results --- p.54 / Chapter 6. --- Conclusions --- p.63 / Chapter 6.1 --- Future development --- p.65 / Appendix A --- p.66 / Determination of the projected path on a mesh: --- p.66 / Reference --- p.68

Typescript (photocopy). / Digitized by Kansas Correctional Industries

Digitized by Kansas Correctional Industries

Typescript (photocopy). / Digitized by Kansas Correctional Industries

This thesis documents a research project that has used computer graphics techniques to reconstruct a set of three-dimensional surfaces from a set of two-dimensional sectional drawings. The work has concentrated on the successful reconstruction of palaeontological specimens, such as brachiopods and early land plants. The reconstruction process is based around a two-stage system. First, the underlying topology of each object is determined automatically using the CorresGrow algorithm, which calculates the correspondence between adjacent sections, many of which have complex contour relationships. Unlike previous solutions, CorresGrow can locate solutions for objects that have multiple, disjoint components. The second stage triangulates the three-dimensional surface using the information provided by the correspondence algorithm. Depending on the similarity in shape of a pair of adjacent contours, the algorithm uses either the original contour vertices or those from the convex hull to perform the surface construction. The other aspects of the project work have looked at the implications of using computer graphics techniques for palaeontological reconstruction. This effects the way in which the data are sampled and digitised so that it is suitable for reconstruction. Using computer graphics also means that the reconstructed models can be used for more than visualisation, in areas such as evolutionary and temporal modelling.

Pang Wai Man. / Thesis submitted in: July 2002. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2003. / Includes bibliographical references (leaves 108-114). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgments --- p.iv / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Image-based Rendering and Modeling --- p.1 / Chapter 1.1.1 --- Image-based versus Geometry-based --- p.5 / Chapter 1.2 --- Capturing for Graphics --- p.6 / Chapter 1.3 --- Organization of this Thesis --- p.8 / Chapter 2 --- Image-based Rendering and Relighting --- p.10 / Chapter 2.1 --- Theoretical Concepts --- p.11 / Chapter 2.1.1 --- Plenoptic Illumination Function --- p.11 / Chapter 2.1.2 --- Apparent BRDF --- p.13 / Chapter 2.1.3 --- Types of lighting --- p.14 / Chapter 2.1.4 --- Image superposition --- p.16 / Chapter 2.2 --- General Rendering Pipeline --- p.18 / Chapter 2.3 --- Rendering Techniques --- p.21 / Chapter 2.3.1 --- Nearest Neighbours and Interpolation --- p.21 / Chapter 2.3.2 --- Image Warping --- p.23 / Chapter 2.4 --- IBR Representations and applications --- p.26 / Chapter 2.4.1 --- Navigation --- p.28 / Chapter 2.4.2 --- Relighting Representations --- p.35 / Chapter 2.4.3 --- High Dynamic Range Imaging --- p.38 / Chapter 2.5 --- Chapter Summary --- p.42 / Chapter 3 --- Capturing Methods --- p.44 / Chapter 3.1 --- Spatial Tracking Approaches --- p.45 / Chapter 3.1.1 --- Mechanical based Method --- p.46 / Chapter 3.1.2 --- Electromagnetic based Method --- p.48 / Chapter 3.1.3 --- Vision based Method --- p.50 / Chapter 3.1.4 --- Comparison --- p.51 / Chapter 3.2 --- High Dynamic Range Imaging --- p.53 / Chapter 3.2.1 --- Successive Exposure Capturing --- p.53 / Chapter 3.2.2 --- Spatial Varing Filter --- p.53 / Chapter 3.2.3 --- Special Designed Hardware --- p.55 / Chapter 3.3 --- Chapter Summary --- p.56 / Chapter 4 --- System Design and Implementation --- p.58 / Chapter 4.1 --- System Overview --- p.58 / Chapter 4.2 --- The Setup --- p.60 / Chapter 4.3 --- Capturing Procedures --- p.61 / Chapter 4.3.1 --- Calibrations --- p.61 / Chapter 4.4 --- Vision based tracking --- p.64 / Chapter 4.4.1 --- The pin-hole camera model --- p.65 / Chapter 4.4.2 --- Basics of Camera Calibration --- p.66 / Chapter 4.5 --- Light Vector Tracking --- p.70 / Chapter 4.5.1 --- The Transformations --- p.70 / Chapter 4.5.2 --- Tracking Accuracy --- p.71 / Chapter 4.5.3 --- Tracking Range Enlargement --- p.72 / Chapter 4.6 --- Capturing Experiment --- p.74 / Chapter 4.7 --- Sampling Analysis --- p.74 / Chapter 4.8 --- Chapter Summary --- p.78 / Chapter 5 --- Data Postprocessing --- p.80 / Chapter 5.1 --- Scattered Data Fitting --- p.81 / Chapter 5.1.1 --- Spherical Delaunay Triangulation --- p.83 / Chapter 5.1.2 --- Interpolation on Sphere --- p.86 / Chapter 5.2 --- Compression --- p.88 / Chapter 5.3 --- Chapter Summary --- p.90 / Chapter 6 --- Relit Results --- p.91 / Chapter 6.1 --- Relighting with Multiple Directional Lights --- p.92 / Chapter 6.2 --- Relighting with Environmental Maps --- p.94 / Chapter 7 --- Conclusion --- p.101 / Chapter 7.1 --- Future Research Aspect --- p.102 / Chapter A --- System User Guide --- p.104 / Chapter A.1 --- Equipment Configuration --- p.104 / Chapter A.2 --- Operation Guide --- p.105 / Chapter A.3 --- Software Components --- p.106 / Chapter A.3.1 --- Image capturing - lightcap --- p.106 / Chapter A.3.2 --- Raw Frame Extraction ´ؤ lfprocess --- p.107 / Chapter A.3.3 --- Resampling and Compression - svscatterppm2urdf . --- p.107 / Bibliography --- p.108

Typescript, etc. / Digitized by Kansas Correctional Industries

No description available.

There are many existing methods for capturing three dimensional data from two dimensional images. Methods based on images captured from multiple view-points require solving the correspondence problem: establishing which points in each image represent the same points in the scene. Most attempts at solving the correspondence problem require carefully controlled lighting and reference points within the scene. A new method captures many consecutive images to form a dense spatiotemporal volume as the camera-or scene-undergoes controlled motion. Feature points in the scene move along predictable paths within this volume. Analysing the exact motion of features determines their three dimensional position in the scene.