The increasing amount of information available in today's world raises the need to retrieve relevant data efficiently. Unlike text-based retrieval, where keywords are successfully used to index into documents, content-based image retrieval poses up front the fundamental questions how to extract useful image features and how to use them for intuitive retrieval. We present a novel approach to the problem of navigating through a collection of images for the purpose of image retrieval, which leads to a new paradigm for image database search. We summarize the appearance of images by distributions of color or texture features, and we define a metric between any two such distributions. This metric, which we call the "Earth Mover's Distance" (EMD), represents the least amount of work that is needed to rearrange the mass is one distribution in order to obtain the other. We show that the EMD matches perceptual dissimilarity better than other dissimilarity measures, and argue that it has many desirable properties for image retrieval. Using this metric, we employ Multi-Dimensional Scaling techniques to embed a group of images as points in a two- or three-dimensional Euclidean space so that their distances reflect image dissimilarities as well as possible. Such geometric embeddings exhibit the structure in the image set at hand, allowing the user to understand better the result of a database query and to refine the query in a perceptually intuitive way.

The increasing amount of information available in today's world raises the need to retrieve relevant data efficiently. Unlike text-based retrieval, where keywords are successfully used to index into documents, content-based image retrieval poses up front the fundamental questions how to extract useful image features and how to use them for intuitive retrieval. We present a novel approach to the problem of navigating through a collection of images for the purpose of image retrieval, which leads to a new paradigm for image database search. We summarize the appearance of images by distributions of color or texture features, and we define a metric between any two such distributions. This metric, which we call the "Earth Mover's Distance" (EMD), represents the least amount of work that is needed to rearrange the mass is one distribution in order to obtain the other. We show that the EMD matches perceptual dissimilarity better than other dissimilarity measures, and argue that it has many desirable properties for image retrieval. Using this metric, we employ Multi-Dimensional Scaling techniques to embed a group of images as points in a two- or three-dimensional Euclidean space so that their distances reflect image dissimilarities as well as possible. Such geometric embeddings exhibit the structure in the image set at hand, allowing the user to understand better the result of a database query and to refine the query in a perceptually intuitive way.

Here are the refereed proceedings of the 5th International Conference on Image and Video Retrieval, CIVR 2006, held in Singapore in July 2006. Presents 18 revised full papers and 30 poster papers, together with extended abstracts of 5 papers of 1 special session and those of 10 demonstration papers. These cover interactive image and video retrieval, semantic image retrieval, visual feature analysis, learning and classification, image and video retrieval metrics, and machine tagging.

Technology has spurred the growth of huge image and video libraries, many growing into the hundreds of terabytes. As a result there is a great demand among organizations for the design of databases that can effectively support the storage, search, retrieval, and transmission of video data. Engineers and researchers in the field demand a comprehensi

Multimedia technologies are rapidly attracting more and more interest every day. The Internet as seen from the end user is one of the reasons for this phenomenon, but not the only one. Video on Demand is one of the buzzwords today, but its real availability to the general public is yet to come. Content providers – such as publishers, broadcasting companies, and audio/video production ?rms – must be able to archive and index their productions for later retrieval. This is a formidable task, even more so when the material to be sorted encompasses many di?erent types of several media and covers a time span of several years. In order for such a vast amount of data to be easily available, existing database design models and indexing methodologies have to be improved and re?ned. In addition, new techniques especially tailored to the various types of multimedia must be devised and evaluated. For archiving and trasmission, data compression is another issue that needs to be addressed. In many cases, it has been found that compression and indexing can be successfully integrated, since compressing the data by ?ltering out irrelevancy implies some degree of und- standing of the content structure.

In modern medicine, imaging is the most effective tool for diagnostics, treatment planning and therapy. Almost all modalities have went to directly digital acquisition techniques and processing of this image data have become an important option for health care in future. This book is written by a team of internationally recognized experts from all over the world. It provides a brief but complete overview on medical image processing and analysis highlighting recent advances that have been made in academics. Color figures are used extensively to illustrate the methods and help the reader to understand the complex topics.

Presently, in our world, visual information dominates. The turn of the millenium marks the age of visual information systems. Enabled by picture sensors of all kinds turning digital, visual information will not only enhance the value of existing information, it will also open up a new horizon of previously untapped information sources. There is a huge demand for visual information access from the consumer. As well, the handling of visual information is boosted by the rapid increase of hardware and Internet capabilities. Advanced technology for visual information systems is more urgently needed than ever before: not only new computational methods to retrieve, index, compress and uncover pictorial information, but also new metaphors to organize user interfaces. Also, new ideas and algorithms are needed which allow access to very large databases of digital pictures and videos. Finally we should not forget new systems with visual interfaces integrating the above components into new types of image, video or multimedia databases and hyperdocuments. All of these technologies will enable the construction of systems that are radically different from conventional information systems. Many novel issues will need to be addressed: query formulation for pictorial information, consistency management thereof, indexing and assessing the quality of these systems. Historically, the expression Visual Information Systems can be understood either as a system for image information or as visual system for any kind information.

This series is directed to diverse managerial professionals who are leading the transformation of individual domains by using expert information and domain knowledge to drive decision support systems (DSSs). The series offers a broad range of subjects addressed in specific areas such as health care, business management, banking, agriculture, environmental improvement, natural resource and spatial management, aviation administration, and hybrid applications of information technology aimed to interdisciplinary issues. This book series is composed of three volumes: Volume 1 consists of general concepts and methodology of DSSs; Volume 2 consists of applications of DSSs in the biomedical domain; Volume 3 consists of hybrid applications of DSSs in multidisciplinary domains. The book is shaped decision support strategies in the new infrastructure that assists the readers in full use of the creative technology to manipulate input data and to transform information into useful decisions for decision makers.

Finally,wearegrateful tooursponsors,theBritishComputerSocietyInformationRetrievalSpecialist Group,theBritishMachineVisionAssociation(BMVA),theInstituteforImage DataResearch,UniversityofNorthumbria,theInstitutionofElectricalEn- neers(IEE),andtheLeidenInstituteofAdvancedComputerScience(LIACS), LeidenUniversiy. May2002 MichaelS. Lew NicuSebe JohnP. Eakins International Conference an Image andVideo Retrieval 2002 Organization Organizing Committee OrganizingCommitteeChair: JohnP. Eakins (UniversityofNorthumbria,UK) TechnicalProgramChair: MichaelS.