Deliverables Distributed query layer ARTISTE has developed an image search and retrieval system that integrates distributed, heterogeneous image collections through use of a 'Distributed Query Layer'. This software: - Transparently translates local metadata schema to common standards such as Dublin Core and CIMI consortium attribute sets to allow cross-collection searching. - Supports a methodology for combining metadata and image content-based analysis into single search queries to enable versatile retrieval and navigation facilities within and between gallery collections. - Provides an open interface for cross-collection search and retrieval that advances existing open standards for remote access to digital libraries, such as OAI (Open Archive Initiative) and ZING SRW (Z39.50 International: Next Generation Search and Retrieval Web Service). A large part of ARTISTE is concerned with use of existing standards for metadata frameworks. However, one area where existing standards have not been sufficient is multimedia content-based search and retrieval. ARTISTE has made a proposal to ZING for additions to SRW. This will enable ARTISTE to make a valued contribution to this rapidly evolving standard. ARTISTE - Database management system ARTISTE is an image search and retrieval system that integrates distributed, heterogeneous image collections. ARTISTE allows users to catalogue art objects and to seamlessly execute content-based and text-based searches over distributed art collections. Key features of ARTISTE include: - Easy to use multilingual web-based access to multiple image collections, using wizards that guide the user in composing and executing complex queries. - Integration of image analysis and text based metadata searches within a single query. - No need to change the local metadata schema of an image collection since ARTISTE can transparently translate to common standards such as Dublin Core and CIMI consortium attribute sets. - Dynamic and automatic linking of search results to external information sources or services, for example reference information on painting techniques or e-commerce sites for gallery merchandise. - Exposure of searching facilities as a search and retrieve web service (SRW) to allow interoperability with distributed 3rd party applications. - Interface to Object Relational Database storage technology to provide a robust and scalable system capable of storing Terabyte data volumes and very high-resolution images. - Sophisticated algorithms that support a wide range of image content-based retrieval searches, for example 'similar colour', 'similar texture' and 'sub image'. Image retrieval and recognition software The image retrieval and recognition software is all the C++ image algorithms and associated tools developed as part of the ARTISTE project. The algorithms include: - RGB/Lab Histogram: Allows retrieval of similar images by example, based on the colour distribution. - Monochrome Histogram: Allows retrieval of similar images by example, based on brightness distribution. - Pyramid Wavelet Transform: Allows retrieval of similar images by example, based on the texture energy distribution. - Multimodal Neighbourhood Search: Allows retrieval of similar images by example, based on the frequency of colour pairs within the image. - Score Normalisation: Allows texture and colour histogram to be used together for possible style searching. - Border finder / classifier: used to detect and classify frames surrounding paintings. - UV Spot Detection: Allows visualisation of areas where the UV reflectance image is black (i.e. No UV reflectance). - Stretcher Detection: Detects stretchers by their line edges so that they can be located, measured, and counted. - Crack Detection/Classification: Detects and classifies cracks from X-Ray images of paintings which show cracks in varnish and paint. - Colour Picker: Allows retrieval of images containing a specified amount of a selected colour or colours. - Query by Fax/Low Quality Image: Allows retrieval of images by example, where the query is a low quality monochrome image such as a fax. This algorithm was also found to be a useful general-purpose matcher in some circumstances. - Colour Coherence Vectors: Allows retrieval of similar images by example, based on the colour distribution with distribution between colours, which are homogenous to some sizeable area. This can be used with sketches, which are generated offline to allow query by sketch. - Grid Based Matching: Allows retrieval of similar images, based on the spatial distribution of features. - Multiscale RGB Histogram: Allows retrieval of similar images by example, based on the colour distribution of images and sub images. - Multiscale Colour Coherence Vectors: Allows retrieval of similar images by example, based on the general colour distribution of the image and sub-images, with discrimination between colours in images which are homogenous to some sizable area. - Multiscale Monochromatic Histogram: Allows retrieval of similar images by example, based on the general distribution of brightness in the image and sub-images. - Multiscale Pyramidal Wavelet Transform: Allows retrieval of similar images by example, based on the general texture distribution of the image and sub-images. - Generic Multiscale Matching: Allows retrieval of similar images by example, based on some feature at various resolutions. Metadata integration for multimedia image collections In ARTISTE, queries can be executed across multiple, distributed image collections without requiring each collection to conform to a standard metadata schema. RDF technology is used to define semantics for standard metadata terms such as Dublin Core and CIMI consortium attribute sets. Each collection provides an RDF mapping that relates these standard metadata terms to individual database table and column values. Queries based on metadata are composed using RDF, and subsequently translated to SQL at each site. RDF is also used to define semantics for combined textual metadata and image content-based querying. The benefits of this approach include: - RDF mapping provides a flexible solution to cross-collection searching. - Mapping to a common schema allows common semantics to be supported without changes to local metadata and schemas. - Users can be dynamically constrained in their querying so that they only request queries that are within the common capabilities of one or more specific collections. - Multilingual translation allows the user to use their native language when specifying which attributes to search over. - Specification of the semantics of both image content and textual metadata based queries allows declarative and seamless integration into a single query language. Dynamic linking layer Traditionally on the WWW, links are hard-coded into HTML pages. In ARTISTE, a separate link database is maintained and the links are applied dynamically at presentation time. Links are applied on a keyword basis. For example, when used on a collection of images and metadata, if the keyword "Van Gogh" is present in the link database, then the link(s) for Van Gogh will be applied every time that the name Van Gogh is displayed to the user. Dynamic linking has several advantages over conventional hard-coded static links: - Links can be easily authored and edited. When links are added or change, modifications can be made to the link database, rather than to all the original documents. - Link presentation can be personalised. For example, 'Van Gogh' might link to Van Gogh books on a gallery's online shop if the user is a member of the general public, or to articles on the restoration of Van Gogh paintings if the user is a restorer. - Multiple links can be applied to a single key word. If there is a choice of links, then the various links can be presented in a table along with a description of each link. Searching for OpenAIRE data... There was an error trying to search data from OpenAIRE No results available
