Wang Yang. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (leaves 80-81). / Abstracts in English and Chinese. / Abstract --- p.i / 摘要 --- p.iii / Table of Contents --- p.iv / List of figures --- p.v / Introduction --- p.1 / Structuring analytical shape grammar --- p.1 / Shape grammar implementation --- p.5 / 3D ting tang generator --- p.7 / Chapter Chapter 1 --- Shape grammar: written grammar and implementations --- p.11 / Shape grammar --- p.11 / Written grammars --- p.12 / Analytical grammars --- p.13 / Shape grammar implementations --- p.14 / Chase's shape grammar system: user interaction mode and emergence --- p.15 / GEdit and Shaper 2D: graphic interface --- p.17 / Chapter Chapter 2 --- Understanding the task of creating ting tang --- p.20 / Model of work --- p.20 / Author of an shape grammar interpreter --- p.21 / What does this interpreter offer? --- p.22 / Before the interpreter: understanding ting tang and the ting tang grammar --- p.23 / The grammar by Li and the grammar for the generator --- p.24 / The process of creating a ting tang in the generator --- p.25 / Chapter Chapter 3 --- The interface and implementation --- p.30 / An illustrated example: --- p.30 / Stage one: --- p.32 / Stage two: --- p.50 / Stage three: install the 3-dimensional components --- p.58 / The two types of step: deterministic and nondeterministic steps --- p.65 / Stage one --- p.65 / Stage two --- p.70 / Stage three --- p.73 / The interface and user control --- p.73 / The back and forward buttons --- p.73 / The inquiry window --- p.74 / Chapter Chapter 4 --- Conclusion --- p.75 / Dynamic rules --- p.75 / Derivation and labels --- p.76 / The limited interface within AutoCAD --- p.77 / Fixed rules --- p.78 / References --- p.80

Tang Yuk Ming = 電影中三維人面模型之估計 / 鄧育明. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2003. / Includes bibliographical references (leaves 107-113). / Text in English; abstracts in English and Chinese. / Tang Yuk Ming = Dian ying zhong san wei ren mian mo xing zhi gu ji / Deng Yuming. / Acknowledgement --- p.i / Abstract --- p.ii / Contents --- p.vi / List of Figures --- p.viii / List of Tables --- p.x / List of Abbreviations and Notations --- p.xi / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Introduction --- p.1 / Chapter 1.2 --- Recent Research Works --- p.2 / Chapter 1.2.1 --- Face modeling from images --- p.2 / Chapter 1.2.2 --- Pose estimation --- p.4 / Chapter 1.3 --- Objectives and Assumptions --- p.7 / Chapter 1.4 --- Our Method --- p.8 / Chapter 1.5 --- Thesis Outline --- p.10 / Chapter 2. --- Basic Theory on 3D Modeling of a Head --- p.11 / Chapter 2.1 --- Introduction --- p.11 / Chapter 2.2 --- Perspective Projection --- p.13 / Chapter 2.3 --- Initialization --- p.17 / Chapter 2.3.1 --- Generic wireframe face model and fiducial points --- p.17 / Chapter 2.3.2 --- Deformations --- p.22 / Chapter 2.3.3 --- Experimental results --- p.35 / Chapter 2.4 --- Summary --- p.39 / Chapter 3. --- Pose Estimation --- p.40 / Chapter 3.1 --- Introduction --- p.40 / Chapter 3.2 --- Problem Description --- p.42 / Chapter 3.3 --- Iterative Least-Square Minimization --- p.45 / Chapter 3.3.1 --- Depth estimation --- p.45 / Chapter 3.3.2 --- Least-square minimization --- p.47 / Chapter 3.3.3 --- Iterative process --- p.52 / Chapter 3.4 --- Experimental Results --- p.54 / Chapter 3.4.1 --- Synthetic data --- p.54 / Chapter 3.4.2 --- Real data --- p.65 / Chapter 3.5 --- Summary --- p.69 / Chapter 4. --- 3D Wireframe Model Estimation --- p.70 / Chapter 4.1 --- Introduction --- p.70 / Chapter 4.2 --- 3D Wireframe Model Estimation --- p.72 / Chapter 4.2.1 --- Least-square minimization --- p.73 / Chapter 4.2.2 --- Iterative process --- p.74 / Chapter 4.3 --- 3D Wireframe Model Estimation of the Subsequent Frames --- p.77 / Chapter 4.4 --- Experimental Results --- p.78 / Chapter 4.4.1 --- Synthetic data --- p.78 / Chapter 4.4.2 --- Real data --- p.84 / Chapter 4.5 --- Summary --- p.98 / Chapter 5. --- Contributions and Conclusions --- p.99 / Chapter 5.1 --- Contributions and conclusions --- p.99 / Chapter 5.2 --- Future Developments --- p.102 / Appendix A Triangles and vertices on the IST model --- p.104 / Bibliography --- p.107

Gong, Boqing. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (p. 53-59). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- 3D Object Representation --- p.1 / Chapter 1.1.1 --- Polygon Mesh --- p.2 / Chapter 1.1.2 --- Voxel --- p.2 / Chapter 1.1.3 --- Range Image --- p.3 / Chapter 1.2 --- Content-Based 3D Object Retrieval --- p.3 / Chapter 1.3 --- 3D Facial Expression Recognition --- p.4 / Chapter 1.4 --- Contributions --- p.5 / Chapter 2 --- 3D Object Retrieval --- p.6 / Chapter 2.1 --- A Conceptual Framework for 3D Object Retrieval --- p.6 / Chapter 2.1.1 --- Query Formulation and User Interface --- p.7 / Chapter 2.1.2 --- Canonical Coordinate Normalization --- p.8 / Chapter 2.1.3 --- Representations of 3D Objects --- p.10 / Chapter 2.1.4 --- Performance Evaluation --- p.11 / Chapter 2.2 --- Public Databases --- p.13 / Chapter 2.2.1 --- Databases of Generic 3D Objects --- p.14 / Chapter 2.2.2 --- A Database of Articulated Objects --- p.15 / Chapter 2.2.3 --- Domain-Specific Databases --- p.15 / Chapter 2.2.4 --- Data Sets for the Shrec Contest --- p.16 / Chapter 2.3 --- Experimental Systems --- p.16 / Chapter 2.4 --- Challenges in 3D Object Retrieval --- p.17 / Chapter 3 --- Boosting 3D Object Retrieval by Object Flexibility --- p.19 / Chapter 3.1 --- Related Work --- p.19 / Chapter 3.2 --- Object Flexibility --- p.21 / Chapter 3.2.1 --- Definition --- p.21 / Chapter 3.2.2 --- Computation of the Flexibility --- p.22 / Chapter 3.3 --- A Flexibility Descriptor for 3D Object Retrieval --- p.24 / Chapter 3.4 --- Enhancing Existing Methods --- p.25 / Chapter 3.5 --- Experiments --- p.26 / Chapter 3.5.1 --- Retrieving Articulated Objects --- p.26 / Chapter 3.5.2 --- Retrieving Generic Objects --- p.27 / Chapter 3.5.3 --- Experiments on Larger Databases --- p.28 / Chapter 3.5.4 --- Comparison of Times for Feature Extraction --- p.31 / Chapter 3.6 --- Conclusions & Analysis --- p.31 / Chapter 4 --- 3D Object Retrieval with Referent Objects --- p.32 / Chapter 4.1 --- 3D Object Retrieval with Prior --- p.32 / Chapter 4.2 --- 3D Object Retrieval with Referent Objects --- p.34 / Chapter 4.2.1 --- Natural and Man-made 3D Object Classification --- p.35 / Chapter 4.2.2 --- Inferring Priors Using 3D Object Classifier --- p.36 / Chapter 4.2.3 --- Reducing False Positives --- p.37 / Chapter 4.3 --- Conclusions and Future Work --- p.38 / Chapter 5 --- 3D Facial Expression Recognition --- p.39 / Chapter 5.1 --- Introduction --- p.39 / Chapter 5.2 --- Separation of BFSC and ESC --- p.43 / Chapter 5.2.1 --- 3D Face Alignment --- p.43 / Chapter 5.2.2 --- Estimation of BFSC --- p.44 / Chapter 5.3 --- Expressional Regions and an Expression Descriptor --- p.45 / Chapter 5.4 --- Experiments --- p.47 / Chapter 5.4.1 --- Testing the Ratio of Preserved Energy in the BFSC Estimation --- p.47 / Chapter 5.4.2 --- Comparison with Related Work --- p.48 / Chapter 5.5 --- Conclusions --- p.50 / Chapter 6 --- Conclusions --- p.51 / Bibliography --- p.53

This thesis explores efficient techniques for generating 3D reconstructions from imagery. Non-linear optimization is one of the core techniques used when computing a reconstruction and is a computational bottleneck for large sets of images. Since non-linear optimization requires a good initialization to avoid getting stuck in local minima, robust systems for generating reconstructions from images build up the reconstruction incrementally. A hierarchical approach is to split up the images into small subsets, reconstruct each subset independently and then hierarchically merge the subsets. Rigidly locking together portions of the reconstructions reduces the number of parameters needed to represent them when merging, thereby lowering the computational cost of the optimization. We present two techniques that involve optimizing with parts of the reconstruction rigidly locked together. In the first, we start by rigidly grouping the cameras and scene features from each of the reconstructions being merged into separate groups. Cameras and scene features are then incrementally unlocked and optimized until the reconstruction is close to the minimum energy. This technique is most effective when the influence of the new measurements is restricted to a small set of parameters. Measurements that stitch together weakly coupled portions of the reconstruction, though, tend to cause deformations in the low error modes of the reconstruction and cannot be efficiently incorporated with the previous technique. To address this, we present a spectral technique for clustering the tightly coupled portions of a reconstruction into rigid groups. Reconstructions partitioned in this manner can closely mimic the poorly conditioned, low error modes, and therefore efficiently incorporate measurements that stitch together weakly coupled portions of the reconstruction. We explain how this technique can be used to scalably and efficiently generate reconstructions from large sets of images.

Vast collections of historical photographs are being digitally archived and placed online, providing an objective record of the last two centuries that remains largely untapped. We propose that time-varying 3D models can pull together and index large collections of images while also serving as a tool of historical discovery, revealing new information about the locations, dates, and contents of historical images. In particular, our goal is to use computer vision techniques to tie together a large set of historical photographs of a given city into a consistent 4D model of the city: a 3D model with time as an additional dimension. To extract 4D city models from historical images, we must perform inference about the position of cameras and scene structure in both space and time. Traditional structure from motion techniques can be used to deal with the spatial problem, while here we focus on the problem of inferring temporal information: a date for each image and a time interval for which each structural element in the scene persists. We first formulate this task as a constraint satisfaction problem based on the visibility of structural elements in each image, resulting in a temporal ordering of images. Next, we present methods to incorporate real date information into the temporal inference solution. Finally, we present a general probabilistic framework for estimating all temporal variables in structure from motion problems, including an unknown date for each camera and an unknown time interval for each structural element. Given a collection of images with mostly unknown or uncertain dates, we can use this framework to automatically recover the dates of all images by reasoning probabilistically about the visibility and existence of objects in the scene. We present results for image collections consisting of hundreds of historical images of cities taken over decades of time, including Manhattan and downtown Atlanta.

Thesis (M.S. in Computer Science)--Naval Postgraduate School, September 2003. / Thesis advisor(s): Don Brutzman, Curtis Blais. Includes bibliographical references (p. 103-104). Also available online.

published_or_final_version / abstract / toc / Computer Science and Information Systems / Master / Master of Philosophy

Optical Confocal Holography is a combination of two well known concepts: confocal microscopy and optical (laser) holography. Confocal microscopy places an aperture at a conjugate focus to the specimen focus. This filters any rays that are not on the focus plane, allowing a 3-dimensional image of the specimen to be built up over a set of planes. Holography is the measurement of both the amplitude and phase characteristics of light. Typically most methods only measure the amplitude of the image. The phenomenon of interference allows the determination of the phase shift for a coherent source as well. The phase information is directly related to the index of refraction of a material, which in turn is a function of the temperature and composition. As a technique, confocal holography holds promise to better characterize many physical processes in materials science, such as combustion and convection. It also may contribute to the biological sciences by imaging low-contrast, weak-phase objects. Thanks to the ongoing, continued improvement in computer processing speed, it has recently become practical to interpret data from confocal holography microscopy with a computer. The objective of the microscope is to non-invasively measure the three-dimensional, internal temperatures and compositions (e.g. solute/solvent gradient) of a specimen. My contributions over the course of two years to the project were: generation and optimization of an optical design with a software package known as Zemax; sourcing and purchasing all components; formation of a CAD model of the microscope; experiments to characterize building vibrations and air currents; and the development of software in Visual Basic to simulate holograms and execute reconstruction algorithms for the specific application of confocal holography.

Thesis (M.C.S.)--Miami University, Dept. of Computer Science and Systems Analysis, 2009. / Title from first page of PDF document. Includes bibliographical references (p. 122-123).

Thesis (Ph. D.)--State University of New York at Binghamton, Thomas J. Watson School of Engineering and Applied Science, Department of Computer Science, 2008. / Includes bibliographical references.