Content-Based Image Retrieval Using Generic Fourier Descriptor and Gabor Filters

Content-based image retrieval (CBIR) is an important research area with applications to a large amount o f image databases and multimedia information. In CBIR, image is described by several low-level image features, such as color, texture, shape or their combination. The focus o f this paper is on the problem o f shape and texture feature extraction and representation for CBIR. Shape is one o f the most important features because it is very important to human perception. Two shape representation methods, region based shape descriptors and contour based shape descriptors have been studied. A new shape descriptor, Generic Fourier Descriptor (GFD) has been discussed and evaluated with existing shape descriptors. Texture is a global feature that could be used to associate related shapes. Gabor filters (GF) has been selected for texture feature extraction. We apply Generic Fourier Descriptor (GFD) for shape feature extraction and Gabor Filters (GF) for texture feature extraction on three different databases, and we successfully combined GFD and GF together for shape and texture feature extraction. Experimental results show that the proposed GFD+GF is robust to all the test databases with best retrieval rate. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACK NO W LEDG EM ENTS This thesis could not be finished without the help and support o f many people who are gratefully acknowledged here. At the very first, I ’m honored to express my deepest gratitude to my dedicated supervisor, Dr. J. Wu; his patient and meticulous guidance and invaluable suggestions are indispensable to the completion o f this thesis. W hat’s more, I wish to extend my thanks to my committee members, for their support o f this study. I owe special thanks to Dr. J. Chen and Dr. H. Wu, for their priceless comments on this study. Thanks are also due to my postgraduate friends, who never failed to give me great encouragement and suggestions. Special thanks should go to Mr. Zhengqiao Ji, Ms. Lan Xu, Miss Kathleen Dizon, and Mr. Tong Wang for their help and support when I had problems writing this thesis. At last but not least, I would like to thank my family for their support all the way from the very beginning o f my postgraduate study. I am thankful to all my family members for their understanding and love. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.