published_or_final_version / abstract / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

In this thesis, we primarily focus on one common type of symmetry, the translational symmetry. We first review the current state-of-the-art methods for translational symmetry detection, and discuss their benefits and drawbacks. Towards an efficient, automatic and widely applicable translational symmetry detector, we develop a novel method for automatically detecting translational symmetry patterns, and extracting the corresponding lattices from images without pre-segmentation or reconstructing the underlying 3D geometry. In particular, we employ a region-based feature and fully utilize its regional properties (shape, orientation and well-defined boundary) to propose the repeated candidates. Compared with traditional treatments, which usually rely on point-based features and group them to propose repeated candidates, our treatment is more efficient and stable to perspective projection, distortion or noise. By clustering the candidate regions and indexing the major clusters using a GPU KD-tree, the parallel lattice formation processes turn out to be very efficient and achieve a real-time rate. By using a set of spatially varying vectors with a loose neighboring constraint to represent the underlying lattice, we successfully detect most of translational symmetry patterns over arbitrary surfaces, which can be planar or curve, without or with perspective projection, and even when suffered from global and local deformations. Moreover, the parallel searching and saving scheme enables us to simultaneously detect multiple disjoint symmetry patterns from an input images. / Symmetry has been an important concept in the nature, science and art. There is an abundant of biological, chemical, and artificial structures captured in many real-world images, exhibiting various forms of symmetry. The symmetry patterns and the repetitive elements reinforce the visual importance and usually make an image more attractive. Although our humans have an excellent innate ability in recognizing symmetry and perceiving its beauty, efficient and automatic symmetry detection from images remains a unsolved challenging problem in computer vision and graphics. Without understanding the high-level semantics of symmetry, editing such images while preserving the repetitions and their relations turns out to be difficult to perform, such as image resizing, image inpainting and image replacement. / The significant improvements of our method in both efficiency and accuracy make it a useful tool from which many applications can benefit. One of them is image resizing. We demonstrate that image resizing can be achieved more effectively if we have a better understanding of the image semantics. By analyzing the translational symmetry patterns, and detecting the underlying lattices in an image, we can summarize, instead of only distorting or cropping, the image content. This opens a new space for image resizing that allows us to manipulate, not only image pixels, but also the semantic cells in the lattice. As a general image contains both symmetry & non-symmetry regions and their natures are different, we propose to resize symmetry regions by summarization and non-symmetry region by optimized warping. In addition, by smoothing the intensity of cells across the lattice, we can further maintain the seamlessness of illumination during the summarization. As the difference in resizing strategy between symmetry regions and non-symmetry region leads to discontinuity at their shared boundary, we propose a framework to minimize the artifact. Experimental results show that, with the high-level knowledge of symmetry, our method outperforms the state-of-the-art resizing techniques. / Wu, Huisi. / Advisers: Tien-Tsin Wong; Pheng-Ann Heng. / Source: Dissertation Abstracts International, Volume: 73-06, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 92-100). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

by Wai Leung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1994. / Includes bibliographical references (leaves 148-150). / Chapter 1. --- Introduction --- p.1 / Chapter 1.1. --- Image Selection/Segmentation --- p.3 / Chapter 1.2. --- Background Image Generation --- p.5 / Chapter 1.3. --- Thesis Organisation --- p.6 / Chapter 2. --- Fundamentals of Digital Image Segmentation --- p.7 / Chapter 2.1. --- Edge-based Segmentation methods --- p.7 / Chapter 2.1.1. --- Edge detection --- p.7 / Chapter 2.1.1.1. --- Gradient operators --- p.9 / Chapter 2.1.1.2. --- Compass operators --- p.10 / Chapter 2.1.1.3. --- Laplace operators and zero crossings --- p.10 / Chapter 2.1.1.4. --- Stochastic gradients --- p.11 / Chapter 2.1.1.5. --- Optimal edge detectors --- p.12 / Chapter 2.1.2. --- Boundary extraction --- p.13 / Chapter 2.1.2.1. --- Contour following --- p.13 / Chapter 2.1.2.2. --- Heuristic graph searching --- p.14 / Chapter 2.1.2.3. --- Dynamic programming --- p.14 / Chapter 2.2. --- Region-based Segmentation Methods --- p.15 / Chapter 2.2.1. --- Thresholding --- p.15 / Chapter 2.2.2. --- Region growing --- p.16 / Chapter 2.2.3. --- Region splitting and merging --- p.17 / Chapter 2.2.4. --- Texture segmentation --- p.19 / Chapter 2.2.4.1. --- Spectral approaches --- p.19 / Chapter 2.2.4.2. --- Statistical methods --- p.21 / Chapter 2.3. --- Works on Colour Image Segmentation --- p.23 / Chapter 3. --- Current Selection Tools for Image Retouching --- p.24 / Chapter 3.1. --- Selection by Region Growing --- p.24 / Chapter 3.2. --- Selection by Edge Finding --- p.26 / Chapter 3.3. --- Some Conclusions --- p.27 / Chapter 4. --- A New Segmentation Tool for Image Retouching --- p.28 / Chapter 4.1. --- Requirement and Development Strategy of the Selection Tool --- p.28 / Chapter 4.2. --- Basic Assumptions --- p.29 / Chapter 4.3. --- Algorithm of the Image Selector --- p.30 / Chapter 4.3.1. --- Boundary representation --- p.30 / Chapter 4.3.2. --- Colour edge detection --- p.31 / Chapter 4.3.2.1. --- Colour gradient --- p.31 / Chapter 4.3.2.2. --- Edge detector --- p.32 / Chapter 4.4. --- Boundary Searching --- p.34 / Chapter 4.4.1. --- The searching algorithm - A* --- p.35 / Chapter 4.4.2. --- The cost function and heuristic for A* algorithm --- p.37 / Chapter 4.4.2.1. --- Pixel cost formulation --- p.38 / Chapter A. --- The reference dependent function y --- p.39 / Chapter 1. --- The three similarity functions --- p.42 / Chapter 2. --- The similarity thresholds --- p.44 / Chapter 3. --- The n-reference dependent function --- p.45 / Chapter B. --- The reference independent function f --- p.46 / Chapter C. --- The actual pixel cost function W --- p.46 / Chapter 4.4.2.2. --- The heuristic function h --- p.47 / Chapter 4.5. --- Implementation --- p.48 / Chapter 4.5.1. --- Work-flow of the image selection tool --- p.48 / Chapter 4.5.2. --- Implementation of the user-input stage --- p.51 / Chapter 4.5.3. --- Implementation of the boundary searching phase --- p.54 / Chapter 4.5.3.1. --- OPEN and CLOSE lists --- p.56 / Chapter 4.5.3.2. --- Tracing back searching path --- p.57 / Chapter 4.5.3.3. --- Edge map --- p.58 / Chapter 4.5.3.4. --- Cost calculation --- p.59 / Chapter 4.5.4. --- Implementation of the boundary connection phase --- p.61 / Chapter 4.6. --- Experiments and Results --- p.63 / Chapter 4.6.1. --- Features exploration --- p.63 / Chapter 4.6.1.1. --- Cost function of the boundary tracer --- p.63 / Chapter 4.6.1.2. --- Reference sensitivity --- p.73 / Chapter 4.6.1.3. --- Boundary connection --- p.75 / Chapter 4.6.2. --- Comparison with current image selection tools --- p.76 / Chapter 4.7. --- Discussion --- p.87 / Chapter 4.8. --- Further Improvement --- p.91 / Chapter 4.8.1. --- A* algorithm --- p.91 / Chapter 4.8.2. --- Colour space --- p.92 / Chapter 4.8.3. --- Improvement in processing speed and quality --- p.93 / Chapter 5. --- Background Image Generation by Image Interpolation --- p.95 / Chapter 5.1. --- Current Filling Tools for Background Image Generation --- p.96 / Chapter 5.1.1. --- The Stamp tool --- p.96 / Chapter 5.1.2. --- The Gradient Fill tool --- p.96 / Chapter 5.2. --- Possible Approaches --- p.98 / Chapter 5.2.1. --- Surface Approximation --- p.98 / Chapter 5.2.2. --- Region Growing and Filling --- p.99 / Chapter 5.3. --- A New Image Interpolation Tool --- p.101 / Chapter 5.3.1. --- Problem analysis and requirement specifications --- p.101 / Chapter 5.3.2. --- The interpolation strategy --- p.103 / Chapter 5.3.3. --- Process overview of the image interpolation --- p.107 / Chapter 5.3.4. --- Data representation --- p.108 / Chapter 5.3.5. --- Implementation --- p.110 / Chapter 5.3.5.1. --- User-input - the first stage --- p.110 / Chapter 5.3.5.2. --- Interpolation preparation - the second stage --- p.112 / Chapter A. --- Boundary list rearrangement --- p.113 / Chapter B. --- Boundary parameterisation --- p.114 / Chapter C. --- Area marking --- p.114 / Chapter D. --- Colour pattern preparation --- p.115 / Chapter 5.3.5.3. --- Interpolation - the third stage --- p.118 / Chapter A. --- Filling curve generation --- p.119 / Chapter 1. --- Linear filling curve generation --- p.119 / Chapter 2. --- Non-linear filling curve generation --- p.120 / Chapter B. --- Colour pattern interpolation --- p.125 / Chapter 5.3.5.4. --- Post-filling - the fourth stage --- p.127 / Chapter 5.3.5.5. --- A working example --- p.128 / Chapter 5.3.6. --- Results --- p.130 / Chapter 5.3.7. --- Discussion --- p.140 / Chapter 5.3.8. --- Limitations and future improvement --- p.141 / Chapter 6. --- Conclusions --- p.144 / Appendix --- p.147 / References --- p.148

by Wong Hin Lau. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1993. / Includes bibliographical references (leaves 118-126).

Woo Hok Luen. / Thesis (M.Sc.)--Chinese University of Hong Kong, 1992. / Includes bibliographical references (leaves 42-43). / Abstract --- p.2 / Chapter Ch. 1 --- Introduction --- p.3 / Chapter Ch. 2 --- Review --- p.4 / Chapter 2.1 --- Definitions and notions --- p.4 / Chapter 2.2 --- Features of a skeleton --- p.5 / Chapter 2.3 --- Parallel and sequential algorithms --- p.6 / Chapter 2.4 --- Distance transformations --- p.7 / Chapter 2.5 --- Relaxation labelling process --- p.8 / Chapter Ch. 3 --- The proposed algorithm --- p.9 / Chapter 3.1 --- Definitions and notions --- p.9 / Chapter 3.2 --- The thinning problem --- p.13 / Chapter 3.3 --- Assigning initial probabilities --- p.14 / Chapter 3.4 --- Iteration schemes --- p.15 / Chapter 3.5 --- Net increment of the skeletal probabilities --- p.16 / Chapter 3.6 --- Net increment of the nonskeletal probabilities --- p.16 / Chapter 3.7 --- Terminating condition --- p.19 / Chapter Ch. 4 --- Experimental results --- p.21 / Chapter 4.1 --- Parameters --- p.21 / Chapter 4.2 --- Results --- p.23 / Chapter Ch. 5 --- Discussion --- p.29 / Chapter 5.1 --- 4-neighbour and 8-neighbour DT --- p.30 / Chapter 5.2 --- 8-connectivity number and safe-point tests --- p.32 / Chapter 5.3 --- Mutual exclusion problem --- p.33 / Chapter 5.4 --- Skeleton not of unit width --- p.34 / Chapter 5.5 --- Development --- p.35 / Chapter Ch. 6 --- Conclusion --- p.35 / Appendix --- p.36 / Reference --- p.42

Lau Hang-yee. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2002. / Includes bibliographical references (leaves 121-123). / Abstracts in English and Chinese. / Abstract in English --- p.i / Abstract in Chinese --- p.iii / List of Figures --- p.ix / List of Tables --- p.xiii / Chapter Chapter 1: --- Introduction --- p.1 / Chapter 1.1 --- Objective --- p.1 / Chapter 1.2 --- Definition ofixel-based Algorithm --- p.2 / Chapter 1.3 --- Definition of Region-based Algorithm --- p.3 / Chapter 1.4 --- Definition of Graph-based Algorithm --- p.5 / Chapter 1.5 --- Colour Spaces --- p.7 / Chapter 1.5.1 --- Basics of Colour Vision --- p.7 / Chapter 1.5.2 --- RGB Model --- p.7 / Chapter 1.5.3 --- HSV Model --- p.8 / Chapter 1.6 --- Organization of Thesis --- p.9 / Chapter Chapter 2: --- 2-Mean Clustering --- p.10 / Chapter 2.1 --- Algorithm --- p.11 / Chapter 2.2 --- Experiment Results --- p.14 / Chapter 2.2.1 --- Intensity Images --- p.14 / Chapter 2.2.2 --- Colour Images (Using RGB) --- p.15 / Chapter 2.2.3 --- Colour Image (Using Hue) --- p.19 / Chapter 2.3 --- Discussions --- p.21 / Chapter 2.3.1 --- Advantages --- p.23 / Chapter 2.3.2 --- Disadvantages --- p.24 / Chapter Chapter 3: --- Region Growing with Region Adjacency Graph --- p.25 / Chapter 3.1 --- Algorithm --- p.26 / Chapter 3.1.1 --- Region Growingrocess --- p.26 / Chapter 3.1.2 --- Region Adjacency Graph Growingrocess --- p.27 / Chapter 3.2 --- Experiment Results --- p.30 / Chapter 3.2.1 --- Intensity Images --- p.30 / Chapter 3.2.2 --- Colour Images (Using RGB) --- p.33 / Chapter 3.2.3 --- Colour Image (Using Hue) --- p.37 / Chapter 3.3 --- Discussions --- p.39 / Chapter 3.3.1 --- Advantages --- p.42 / Chapter 3.3.2 --- Disadvantages --- p.43 / Chapter Chapter 4: --- Normalized Cuts --- p.45 / Chapter 4.1 --- Formulation of the Generalized Eigenvaluesystem --- p.47 / Chapter 4.2 --- Algorithm --- p.52 / Chapter 4.3 --- Modification of NC --- p.54 / Chapter 4.4 --- Experiment Results --- p.55 / Chapter 4.4.1 --- Feature Images --- p.56 / Chapter 4.4.2 --- Intensity Images --- p.64 / Chapter 4.4.3 --- Colour Images --- p.66 / Chapter 4.5 --- Discussions --- p.70 / Chapter 4.5.1 --- Advantages --- p.70 / Chapter 4.5.2 --- Disadvantages --- p.71 / Chapter Chapter 5: --- ixel-Region-Graph Method --- p.72 / Chapter 5.1 --- Algorithm --- p.74 / Chapter 5.1.1 --- Step --- p.75 / Chapter 5.1.2 --- Step R --- p.76 / Chapter 5.1.3 --- Step G --- p.77 / Chapter 5.2 --- CompareRG Method with Other Methods --- p.80 / Chapter 5.3 --- Experiment Results --- p.82 / Chapter 5.3.1 --- Feature Image --- p.82 / Chapter 5.3.2 --- Real Images - Set 1 --- p.84 / Chapter 5.3.3 --- Real Images - Set 2 --- p.87 / Chapter 5.3.4 --- Real Images - Set 3 --- p.94 / Chapter 5.3.5 --- Real Images - Set 4 --- p.96 / Chapter 5.4 --- Discussion --- p.97 / Chapter 5.4.1 --- Advantages --- p.99 / Chapter 5.4.2 --- Disadvantages --- p.99 / Chapter 5.5 --- Image Graph --- p.100 / Chapter Chapter 6: --- Conclusion --- p.103 / Appendix I --- p.107 / Appendix II --- p.120 / References --- p.121

我们提出了两个新的方法来解决低级别计算机视觉任务，即图像共同分割和降噪。 / 在共同分割模型上，我们发现对象对应可以为前景统计估计提供有用的信息。我们的方法可以处理极具挑战性的场景，如变形，角度的变化和显着不同的视角和尺度。此外，我们研究了一种新的能量最小化模型，可以同时处理多个图像。真实和基准数据的定性和定量实验证明该方法的有效性。 / 另一方面，噪音始终和高频图像结构是紧耦合的，从而使得减少噪音非常很难。在我们的降噪模型中，我们建议稍微使图像光学离焦，以减少图像和噪声的耦合。这使得我们能更有效地降低噪音，随后恢复失焦。我们的分析显示，这是可能的，并且用许多例子证明我们的技术，其中包括低光图像。 / We present two novel methods to tackle low level computer vision tasks,i.e., image cosegmentation and denoise . / In our cosegmentationmodel, we discover object correspondence canprovide useful information for foreground statistical estimation. Ourmethod can handle extremely challenging scenarios such as deformation, perspective changes and dramatically different viewpoints/scales. In addition, we develop a novel energy minimization model that can handlemultiple images. Experiments on real and benchmark data qualitatively and quantitatively demonstrate the effectiveness of the approach. / One the other hand, noise is always tightly coupled with high-frequencyimage structure, making noise reduction generally very difficult. In ourdenoise model, we propose slightly optically defocusing the image in orderto loosen this noise-image structure coupling. This allows us to more effectively reduce noise and subsequently restore the small defocus. Weanalytically show how this is possible, and demonstrate our technique on a number of examples that include low-light images. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Qin, Zenglu. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 64-71). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese. / Abstract --- p.i / Acknowledgement --- p.ii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Motivation and Objectives --- p.1 / Chapter 1.1.1 --- Cosegmentation --- p.1 / Chapter 1.1.2 --- Image Denoise --- p.4 / Chapter 1.2 --- Thesis Outline --- p.7 / Chapter 2 --- Background --- p.8 / Chapter 2.1 --- Cosegmentation --- p.8 / Chapter 2.2 --- Image Denoise --- p.10 / Chapter 3 --- Cosegmentation of Multiple Deformable Objects --- p.12 / Chapter 3.1 --- Related Work --- p.12 / Chapter 3.2 --- Object Corresponding Cosegmentation --- p.13 / Chapter 3.3 --- Importance Map with Object Correspondence --- p.15 / Chapter 3.3.1 --- Feature Importance Map --- p.16 / Chapter 3.3.2 --- Importance Energy E[subscript i](xp) --- p.20 / Chapter 3.4 --- Experimental Result --- p.20 / Chapter 3.4.1 --- Two-Image Cosegmentation --- p.21 / Chapter 3.4.2 --- ETHZ Toys Dataset --- p.22 / Chapter 3.4.3 --- More Results --- p.24 / Chapter 3.5 --- Summary --- p.27 / Chapter 4 --- Using Optical Defocus to Denoise --- p.28 / Chapter 4.1 --- Related Work --- p.29 / Chapter 4.2 --- Noise Analysis --- p.30 / Chapter 4.3 --- Noise Estimation with Focal Blur --- p.33 / Chapter 4.3.1 --- Noise Estimation with a Convolution Model --- p.33 / Chapter 4.3.2 --- Determining λ --- p.41 / Chapter 4.4 --- Final Deconvolution and Error Analysis --- p.43 / Chapter 4.5 --- Implementation --- p.45 / Chapter 4.6 --- Quantitative Evaluation --- p.47 / Chapter 4.7 --- More Experimental Results --- p.53 / Chapter 4.8 --- Summary --- p.56 / Chapter 5 --- Conclusion --- p.62 / Bibliography --- p.64

Typescript (photocopy). / Digitized by Kansas Correctional Industries