Epic A photograph retrieval system based on evidence combination approach.pdf

下载文档

4
0
约4.61万字
约 10页
2017-04-11 发布于江苏
举报
版权申诉
保障服务

Epic A photograph retrieval system based on evidence combination approach.pdf

1、本文档共10页，可阅读全部内容。
2、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

Epic A photograph retrieval system based on evidence combination approach

Epic: A Photograph Retrieval System Based on Evidence CombinationApproachJoemon M. Jose and David J. HarperThe Robert Gordon UniversityAberdeen, UK, AB25 1HG.fj.jose,d.harperg@scms.rgu.ac.uk. AbstractIn this paper, we describe a system which applies an evi-dence combination approach for the retrieval of photographsfrom large repositories. To achieve eective retrieval we usefeatures derived from various components (such as text, andimage) of the photographic document. As photographic fea-tures we use spatial features, which are objects and theirlocations in the photograph. We use these spatial featuresalong with other textual features derived from the variouscomponents of the document. A user can search based onany of the features or in combination. For the combinationof evidence we apply the Dempster-Shafer theory. In thiscombination scheme, we achieve the integration of databaseand information retrieval mechanisms into one computa-tional model.The architecture of the system is based on a client-serverarchitecture. The server is built using ECLAIR, an extensi-ble class library for information retrieval developed on topof an object-oriented database management system. Theclient is built using the Java programming language. Thearchitecture of the server allows the integration of DB andIR approaches at the implementation level. The usefulnessof the spatial features is evaluated using practicing graphicdesigners and has given very promising results.KEYWORDS: Photograph Information Retrieval, Mul-timedia Databases, Database and Information Retrieval In-tegration, Spatial Features, Evidence Combination1 IntroductionDue to the widespread use of photographs in everyday life,there is a need for storing, managing and retrieving pho-tographs. A large amount of pictorial information is pro-duced in domains like the newspaper industry, the advertis-ing business and the publishing industry. Journalists maylook for photographs that describe an event or concern aperson. Advertisin