• Login
    About WHOAS
    View Item 
    •   WHOAS Home
    • Woods Hole Oceanographic Institution
    • Academic Programs
    • WHOI Theses
    • View Item
    •   WHOAS Home
    • Woods Hole Oceanographic Institution
    • Academic Programs
    • WHOI Theses
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Browse

    All of WHOASCommunities & CollectionsBy Issue DateAuthorsTitlesKeywordsThis CollectionBy Issue DateAuthorsTitlesKeywords

    My Account

    LoginRegister

    Statistics

    View Usage Statistics

    Transform texture classification

    Thumbnail
    View/Open
    Tang_Thesis (5.538Mb)
    Date
    1996-05
    Author
    Tang, Xiaoou  Concept link
    Metadata
    Show full item record
    Citable URI
    https://hdl.handle.net/1912/5690
    DOI
    10.1575/1912/5690
    Keyword
     Signal processing; Plankton populations 
    Abstract
    This thesis addresses the three major components of a texture classification system: texture image transform, feature extraction/selection, and classification. A unique theoretical investigation of texture analysis, drawing on an extensive survey of existing approaches, defines the interrelations among 11 types of texture analysis methods. A novel unification of the different methods defines a framework of transformation and representation in which three major classes of transform matrices capture texture information of increasing coherence length or correlation distance: the spatial domain method (co-occurrence method), the micro-structural method (run-length method), and the frequency multi-channel method (Fourier spectrum method). A more concise vector representation of a selected transform matrix is then needed for input to a classifier. Unlike traditional methods, which use various special functions to describe the properties of each transform matrix, a new approach directly applies a principle component analysis technique to the transform matrix. The Karhunen-Loeve Transfonn (KLT) extracts a vector of dominant features, optimally preserving texture information in the matrix. This approach is made possible by the introduction of a novel Multi-level Dominant Eigenvector Estimation (MDEE) algorithm, which reduces the computational complexity of the standard KLT by several orders of magnitude. The statistical Bhattacharyya distance measure is then used to rank dominant features according to their discrimination power. Experimental results of applying the new algorithm to the three transform matrix classes show a strong increase in performance by texture analysis methods traditionally considered to be least efficient For example, the power spectrum and run-length methods now rank among the best. Using the same MDEE algorithm, the three extracted feature vectors are then combined into a more complete description of texture images. The same approach is also used for a study of object recognition, where the combined vector also include granulometric; object-boundary, and moment-invariant features. In most classification experiments, a simple statistical Gaussian classifier is used. The plankton object recognition experiments use a Learning Vector Quantization (LVQ) neural-net classifier to achieve superior performance on the highly non-uniform plankton database. By introducing a new parallel LVQ learning scheme, the speed of network training is dramatically increased. Tests show a 95% classification accuracy on six plankton taxa taken from nearly 2,000 images. This result is comparable with what a trained biologist can accomplish by traditional manual techniques, making possible for the first time a fully automated, at-sea approach to real-time mapping of plankton populations.
    Description
    Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution May 1996
    Collections
    • Applied Ocean Physics and Engineering (AOP&E)
    • WHOI Theses
    Suggested Citation
    Thesis: Tang, Xiaoou, "Transform texture classification", 1996-05, DOI:10.1575/1912/5690, https://hdl.handle.net/1912/5690
     

    Related items

    Showing items related by title, author, creator and subject.

    • Thumbnail

      The spatial and temporal abundance patterns of chaetognaths in the western North Atlantic Ocean 

      Cheney, Jerry (Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 1982-08)
      Zooplankton samples were collected with the MOCNESS (Multiple Opening/ Closing Net and Environmental Sensing System) on six cruises in the western North Atlantic Ocean during the period from August 1975 to November 1977 ...
    • Thumbnail

      Mantle upwelling, melt generation, and magma transport beneath mid-ocean ridges 

      Magde, Laura S. (Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 1997-03)
      The formation of new oceanic crust is the result of a complex geodynamic system in which mantle rises beneath spreading centers and undergoes decompression melting. The melt segregates from the matrix and is focused to ...
    • Thumbnail

      Experimental and seismological constraints on the rheology, evolution, and alteration of the lithosphere at oceanic spreading centers 

      deMartin, Brian J. (Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 2007-02)
      Oceanic spreading centers are sites of magmatic, tectonic, and hydrothermal processes. In this thesis I present experimental and seismological constraints on the evolution of these complex regions of focused crustal ...
    All Items in WHOAS are protected by original copyright, with all rights reserved, unless otherwise indicated. WHOAS also supports the use of the Creative Commons licenses for original content.
    A service of the MBLWHOI Library | About WHOAS
    Contact Us | Send Feedback | Privacy Policy
    Core Trust Logo