Search (2 results, page 1 of 1)

0.11656364 = sum of:
  0.078996494 = product of:
    0.23698948 = sum of:
      0.23698948 = weight(_text_:objects in 4253) [ClassicSimilarity], result of:
        0.23698948 = score(doc=4253,freq=20.0), product of:
          0.31904724 = queryWeight, product of:
            5.315071 = idf(docFreq=590, maxDocs=44218)
            0.060026903 = queryNorm
          0.74280375 = fieldWeight in 4253, product of:
            4.472136 = tf(freq=20.0), with freq of:
              20.0 = termFreq=20.0
            5.315071 = idf(docFreq=590, maxDocs=44218)
            0.03125 = fieldNorm(doc=4253)
    0.33333334 = coord(1/3)
  0.037567146 = product of:
    0.07513429 = sum of:
      0.07513429 = weight(_text_:encyclopedia in 4253) [ClassicSimilarity], result of:
        0.07513429 = score(doc=4253,freq=2.0), product of:
          0.3194549 = queryWeight, product of:
            5.321862 = idf(docFreq=586, maxDocs=44218)
            0.060026903 = queryNorm
          0.2351953 = fieldWeight in 4253, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            5.321862 = idf(docFreq=586, maxDocs=44218)
            0.03125 = fieldNorm(doc=4253)
    0.5 = coord(1/2)

Abstract

In this chapter we describe methodologies for representing information in digital libraries in order to support efficient search and content-based retrieval, and we focus our attention an image repositories. We identify different abstraction levels at which content can be specified. We discuss how simple objects can be defined as connected regions that are homogeneous with respect to pixel-level, feature-level, semantic-level, and metadata-level characteristics. We describe how information can be efficiently represented at these different levels, how a user can specify content, and what mechanisms can be used to perform the search. We present techniques for combining image representation, in particular image compression, with image-processing operators designed for content-based searches. On the one hand this approach makes it possible to extract and index content at database ingestion time even when the data volume is large. On the other hand it allows the system to retrieve simple objects for which definitions are provided at query time and that have not been pre-extracted and preindexed. Simple objects, however, are not sufficient to describe the richness of image content. We rely an the concept of composite object, a collection of simple objects satisfying a set of relations, to specify complex queries. We describe algorithms for retrieving composite objects from image databases. This article is organized as follows. The next section contains an introduction to digital libraries and a description of four application areas. The aspects of compression that are relevant to progressive retrieval are discussed in the section attached. The subsequent section provides a brief introduction to content-based searches, the standard approach to query specification in multimedia databases. Our definition of content in terms of simple and composite objects is contained in the following section. Simple objects can be defined at multiple abstraction levels; fundamental concepts, content-extraction methodologies including progressive techniques, and indexing methods are discussed in the next section. Simple objects can also be defined simultaneously at multiple abstraction levels, and aggregated to form composite objects. The semantics of both types of objects and the techniques required to search for them are the subject of the section after that. The final section contains the conclusions.

Source

Encyclopedia of library and information science. Vol.71, [=Suppl.34]

0.10945906 = sum of:
  0.04371655 = product of:
    0.13114965 = sum of:
      0.13114965 = weight(_text_:objects in 3885) [ClassicSimilarity], result of:
        0.13114965 = score(doc=3885,freq=2.0), product of:
          0.31904724 = queryWeight, product of:
            5.315071 = idf(docFreq=590, maxDocs=44218)
            0.060026903 = queryNorm
          0.41106653 = fieldWeight in 3885, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            5.315071 = idf(docFreq=590, maxDocs=44218)
            0.0546875 = fieldNorm(doc=3885)
    0.33333334 = coord(1/3)
  0.06574251 = product of:
    0.13148502 = sum of:
      0.13148502 = weight(_text_:encyclopedia in 3885) [ClassicSimilarity], result of:
        0.13148502 = score(doc=3885,freq=2.0), product of:
          0.3194549 = queryWeight, product of:
            5.321862 = idf(docFreq=586, maxDocs=44218)
            0.060026903 = queryNorm
          0.41159177 = fieldWeight in 3885, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            5.321862 = idf(docFreq=586, maxDocs=44218)
            0.0546875 = fieldNorm(doc=3885)
    0.5 = coord(1/2)

Abstract

We describe approaches and techniques for indexing and retrieving still images from multimedia databases. We specifically emphasize content-based image retrieval (CBIR), a class of techniques where the user composes queries that specify the content of the desired images. After a brief overview of digital image formats, we analyze different approaches to content specification: in terms of low-level visual features, of objects, and of metadata. We then describe a general progressive framework that combines these approaches. We finally conclude the entry with an overview of common applications of image repositories and digital libraries, such as medical imaging, remote-sensing imaging, and data for the oil industry.

Source

Encyclopedia of library and information sciences. 3rd ed. Ed.: M.J. Bates