| In various application domains such as entertainment, biomedicine, commerce, education, and crime prevention, the volume of digital data archives is growing rapidly. The very large repository of digital information raises challenging problems in retrieval and various other information manipulation tasks. Content-based image retrieval (CBIR) is aimed at efficient retrieval of relevant images from large image databases based on automatically derived imagery features. However, images with high feature similarities to the query image may be very different from the query in terms of semantics. This discrepancy between low-level content features (such as color, texture, and shape) and high-level semantic concepts (such as sunset, flowers, outdoor scene, etc.) is known as “semantic gap,” which is an open challenging problem in current CBIR systems.; With the ultimate goal of narrowing the semantic gap, this thesis makes three contributions to the field of CBIR. The first contribution is a novel region-based image similarity measure. An image is represented by a set of segmented regions each of which is characterized by a fuzzy feature (fuzzy set) reflecting color, texture, and shape properties. Fuzzy features naturally characterize the gradual transition between regions (blurry boundaries) within an image, and incorporate the segmentation-related uncertainties into the retrieval algorithm. The resemblance of two images is defined as the overall similarity between two families of fuzzy features, and quantified by a similarity measure that integrates properties of all the regions in the images. Compared with similarity measures based on individual regions and on all regions with crisp-valued feature representations, the proposed measure greatly reduces the influence of inaccurate segmentation, and provides a very intuitive quantification.; The second contribution is a novel image retrieval scheme using unsupervised learning. It is built on a hypothesis that images of the same semantics tend to be clustered in some feature space. The proposed method attempts to capture semantic concepts by learning the way that images of the same semantics are similar and retrieving image clusters instead of a set of ordered images. Clustering is dynamic. In particular, clusters formed depend on which images are retrieved in response to the query. Therefore, the clusters give the algorithm as well as the users semantic relevant clues as to where to navigate. The proposed retrieval scheme is a general approach that can be combined with any real-valued symmetric similarity measure (metric or nonmetric). Thus it may be embedded in many current CBIR systems.; The last contribution is a novel region-based image classification method. An image is represented as a set of regions obtained from image segmentation. It is assumed that the concept underlying an image category is related to the occurrence of regions of certain types, which are called region prototypes (RPs), in an image. Each RP represents a class of regions that is more likely to appear in images with the specific label than in the other images, and is found according to an objective function measuring a co-occurrence of similar regions from different images with the same label. An image classifier is then defined by a set of rules associating the appearance of RPs in an image with image labels. The learning of such classifiers is formulated as a Support Vector Machine (SVM) learning problem with a special class of kernels. |
