English  |  正體中文  |  简体中文  |  全文筆數/總筆數 : 49378/84106 (59%)
造訪人次 : 7377765      線上人數 : 69
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
搜尋範圍 查詢小技巧:
  • 您可在西文檢索詞彙前後加上"雙引號",以獲取較精準的檢索結果
  • 若欲以作者姓名搜尋,建議至進階搜尋限定作者欄位,可獲得較完整資料
  • 進階搜尋
    請使用永久網址來引用或連結此文件: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/45327

    題名: Image indexing and similarity retrieval based on spatial relationship model
    作者: 王英宏;Wang, Ying-hong
    貢獻者: 淡江大學資訊工程學系
    關鍵詞: Image retrieval;Image database;Spatial knowledge;Spatial reasoning;Similarity retrieval;2D Strings;LCS algorithm;2D Be-string
    日期: 2003-08
    上傳時間: 2010-03-26 18:59:18 (UTC+8)
    出版者: Elsevier
    摘要: The increasing availability of image and multimedia-oriented applications markedly impacts image/multimedia file and database systems. Image data are not well-defined keywords such as traditional text data used in searching and retrieving functions. Consequently, various indexing and retrieving methodologies must be defined based on the characteristics of image data. Spatial relationships represent an important feature of objects (called icons) in an image (or picture). Spatial representation by 2D String and its variants, in a pictorial spatial database, has been attracting growing interest. However, most 2D Strings represent spatial information by cutting the icons out of an image and associating them with many spatial operators. The similarity retrievals by 2D Strings require massive geometric computation and focus only on those database images that have all the icons and spatial relationships of the query image. This study proposes a new spatial-relationship representation model called “two dimension begin-end boundary string” (2D Be-string). The 2D Be-string represents an icon by its MBR boundaries. By applying “dummy objects”, the 2D Be-string can intuitively and naturally represent the pictorial spatial information without any spatial operator. A method of evaluating image similarities, based on the modified “longest common subsequence” algorithm, is presented. The proposed evaluation method cannot only sift out those images of which all icons and their spatial relationships fully accord with query images, but for those images some of whose icons and/or spatial relationships are similar to those of query images. Problems of uncertainty the query targets and/or spatial relationships thus solved. The representation model and similarity evaluation also simplify the retrieval progress of linear transformations, including rotation and reflection, of images.
    關聯: Information sciences 154(1-2), pp.39-58
    DOI: 10.1016/S0020-0255(03)00005-7
    顯示於類別:[資訊工程學系暨研究所] 期刊論文


    檔案 大小格式瀏覽次數



    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - 回饋