published_or_final_version / Computer Science / Doctoral / Doctor of Philosophy

by Chan Yuk Ming. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1999. / Includes bibliographical references (leaves 117-124). / Abstract also in Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Content-based Image Retrieval --- p.3 / Chapter 1.2 --- Designing a Shape-based Image Retrieval System --- p.4 / Chapter 1.3 --- Information on Trademark --- p.6 / Chapter 1.3.1 --- What is a Trademark? --- p.6 / Chapter 1.3.2 --- Search for Conflicting Trademarks --- p.7 / Chapter 1.3.3 --- Research Scope --- p.8 / Chapter 1.4 --- Information on Chinese Cursive Script Character --- p.9 / Chapter 1.5 --- Problem Definition --- p.9 / Chapter 1.6 --- Contributions --- p.11 / Chapter 1.7 --- Thesis Organization --- p.13 / Chapter 2 --- Literature Review --- p.14 / Chapter 2.1 --- Trademark Retrieval using QBIC Technology --- p.14 / Chapter 2.2 --- STAR --- p.16 / Chapter 2.3 --- ARTISAN --- p.17 / Chapter 2.4 --- Trademark Retrieval using a Visually Salient Feature --- p.18 / Chapter 2.5 --- Trademark Recognition using Closed Contours --- p.19 / Chapter 2.6 --- Trademark Retrieval using a Two Stage Hierarchy --- p.19 / Chapter 2.7 --- Logo Matching using Negative Shape Features --- p.21 / Chapter 2.8 --- Chapter Summary --- p.22 / Chapter 3 --- Background on Shape Representation and Matching --- p.24 / Chapter 3.1 --- Simple Geometric Features --- p.25 / Chapter 3.1.1 --- Circularity --- p.25 / Chapter 3.1.2 --- Rectangularity --- p.26 / Chapter 3.1.3 --- Hole Area Ratio --- p.27 / Chapter 3.1.4 --- Horizontal Gap Ratio --- p.27 / Chapter 3.1.5 --- Vertical Gap Ratio --- p.28 / Chapter 3.1.6 --- Central Moments --- p.28 / Chapter 3.1.7 --- Major Axis Orientation --- p.29 / Chapter 3.1.8 --- Eccentricity --- p.30 / Chapter 3.2 --- Fourier Descriptors --- p.30 / Chapter 3.3 --- Chain Codes --- p.31 / Chapter 3.4 --- Seven Invariant Moments --- p.33 / Chapter 3.5 --- Zernike Moments --- p.35 / Chapter 3.6 --- Edge Direction Histogram --- p.36 / Chapter 3.7 --- Curvature Scale Space Representation --- p.37 / Chapter 3.8 --- Chapter Summary --- p.39 / Chapter 4 --- Genetic Algorithm for Weight Assignment --- p.42 / Chapter 4.1 --- Genetic Algorithm (GA) --- p.42 / Chapter 4.1.1 --- Basic Idea --- p.43 / Chapter 4.1.2 --- Genetic Operators --- p.44 / Chapter 4.2 --- Why GA? --- p.45 / Chapter 4.3 --- Weight Assignment Problem --- p.46 / Chapter 4.3.1 --- Integration of Image Attributes --- p.46 / Chapter 4.4 --- Proposed Solution --- p.47 / Chapter 4.4.1 --- Formalization --- p.47 / Chapter 4.4.2 --- Proposed Genetic Algorithm --- p.43 / Chapter 4.5 --- Chapter Summary --- p.49 / Chapter 5 --- Shape-based Trademark Image Retrieval System --- p.50 / Chapter 5.1 --- Problems on Existing Methods --- p.50 / Chapter 5.1.1 --- Edge Direction Histogram --- p.51 / Chapter 5.1.2 --- Boundary Based Techniques --- p.52 / Chapter 5.2 --- Proposed Solution --- p.53 / Chapter 5.2.1 --- Image Preprocessing --- p.53 / Chapter 5.2.2 --- Automatic Feature Extraction --- p.54 / Chapter 5.2.3 --- Approximated Boundary --- p.55 / Chapter 5.2.4 --- Integration of Shape Features and Query Processing --- p.58 / Chapter 5.3 --- Experimental Results --- p.58 / Chapter 5.3.1 --- Experiment 1: Weight Assignment using Genetic Algorithm --- p.59 / Chapter 5.3.2 --- Experiment 2: Speed on Feature Extraction and Retrieval --- p.62 / Chapter 5.3.3 --- Experiment 3: Evaluation by Precision --- p.63 / Chapter 5.3.4 --- Experiment 4: Evaluation by Recall for Deformed Images --- p.64 / Chapter 5.3.5 --- Experiment 5: Evaluation by Recall for Hand Drawn Query Trademarks --- p.66 / Chapter 5.3.6 --- "Experiment 6: Evaluation by Recall for Rotated, Scaled and Mirrored Images" --- p.66 / Chapter 5.3.7 --- Experiment 7: Comparison of Different Integration Methods --- p.68 / Chapter 5.4 --- Chapter Summary --- p.71 / Chapter 6 --- Shape-based Chinese Cursive Script Character Image Retrieval System --- p.72 / Chapter 6.1 --- Comparison to Trademark Retrieval Problem --- p.79 / Chapter 6.1.1 --- Feature Selection --- p.73 / Chapter 6.1.2 --- Speed of System --- p.73 / Chapter 6.1.3 --- Variation of Style --- p.73 / Chapter 6.2 --- Target of the Research --- p.74 / Chapter 6.3 --- Proposed Solution --- p.75 / Chapter 6.3.1 --- Image Preprocessing --- p.75 / Chapter 6.3.2 --- Automatic Feature Extraction --- p.76 / Chapter 6.3.3 --- Thinned Image and Linearly Normalized Image --- p.76 / Chapter 6.3.4 --- Edge Directions --- p.77 / Chapter 6.3.5 --- Integration of Shape Features --- p.78 / Chapter 6.4 --- Experimental Results --- p.79 / Chapter 6.4.1 --- Experiment 8: Weight Assignment using Genetic Algorithm --- p.79 / Chapter 6.4.2 --- Experiment 9: Speed on Feature Extraction and Retrieval --- p.81 / Chapter 6.4.3 --- Experiment 10: Evaluation by Recall for Deformed Images --- p.82 / Chapter 6.4.4 --- Experiment 11: Evaluation by Recall for Rotated and Scaled Images --- p.83 / Chapter 6.4.5 --- Experiment 12: Comparison of Different Integration Methods --- p.85 / Chapter 6.5 --- Chapter Summary --- p.87 / Chapter 7 --- Conclusion --- p.88 / Chapter 7.1 --- Summary --- p.88 / Chapter 7.2 --- Future Research --- p.89 / Chapter 7.2.1 --- Limitations --- p.89 / Chapter 7.2.2 --- Future Directions --- p.90 / Chapter A --- A Representative Subset of Trademark Images --- p.91 / Chapter B --- A Representative Subset of Cursive Script Character Images --- p.93 / Chapter C --- Shape Feature Extraction Toolbox for Matlab V53 --- p.95 / Chapter C.l --- central .moment --- p.95 / Chapter C.2 --- centroid --- p.96 / Chapter C.3 --- cir --- p.96 / Chapter C.4 --- ess --- p.97 / Chapter C.5 --- css_match --- p.100 / Chapter C.6 --- ecc --- p.102 / Chapter C.7 --- edge一directions --- p.102 / Chapter C.8 --- fourier-d --- p.105 / Chapter C.9 --- gen_shape --- p.106 / Chapter C.10 --- hu7 --- p.108 / Chapter C.11 --- isclockwise --- p.109 / Chapter C.12 --- moment --- p.110 / Chapter C.13 --- normalized-moment --- p.111 / Chapter C.14 --- orientation --- p.111 / Chapter C.15 --- resample-pts --- p.112 / Chapter C.16 --- rectangularity --- p.113 / Chapter C.17 --- trace-points --- p.114 / Chapter C.18 --- warp-conv --- p.115 / Bibliography --- p.117

Mo, Deyuan. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 153-167). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese; appendixes includes Chinese.

University of Macau / Faculty of Science and Technology. / Department of Computer and Information Science