Recognition and assessment of seafloor vegetation using a single beam echosounder

Created by W.Langdon from gp-bibliography.bib Revision:1.3973

@PhdThesis{Yao-Ting.Tseng:thesis,
  title =        "Recognition and assessment of seafloor vegetation
                 using a single beam echosounder",
  author =       "Yao-Ting Tseng",
  year =         "2009",
  school =       "Department of Imaging and Applied Physics, Centre for
                 Marine Science and Technology, Curtin University of
                 Technology",
  address =      "Perth, Australia",
  month =        feb,
  keywords =     "genetic algorithms, genetic programming, GPLAB, single
                 beam echo sounder, sea floor vegetation, recognition
                 and assessment, plant benthos, ecological balance,
                 environmental managers",
  bibsource =    "OAI-PMH server at espace.library.curtin.edu.au",
  language =     "en",
  oai =          "oai:espace.library.curtin.edu.au:128517",
  rights =       "unrestricted",
  URL =          "http://espace.library.curtin.edu.au:80/R?func=dbin-jump-full&local_base=gen01-era02&object_id=128517",
  size =         "246 pages",
  abstract =     "This study focuses on the potential of using a single
                 beam echosounder as a tool for recognition and
                 assessment of seafloor vegetation. Sea floor vegetation
                 is plant benthos and occupies a large portion of the
                 shallow coastal bottoms. It plays a key role in
                 maintaining the ecological balance by influencing the
                 marine and terrestrial worlds through interactions with
                 its surrounding environment. Understanding of its
                 existence on the sea floor is essential for
                 environmental managers.

                 Due to the important role of sea floor vegetation to
                 the environment, a detailed investigation of acoustic
                 methods that can provide effective recognition and
                 assessment of the sea floor vegetation by using
                 available sonar systems is necessary. One of the
                 frequently adopted approaches to the understanding of
                 ocean environment is through the mapping of the sea
                 floor. Available acoustic techniques vary in kinds and
                 are used for different purposes. Because of the wide
                 scope of available techniques and methods which can be
                 employed in the field, this study has limited itself to
                 sonar techniques of normal incidence configuration
                 relative to sea-floors in selected regions and for
                 particular marine habitats. For this study, a single
                 beam echo-sounder operating at two frequencies was
                 employed. Integrated with the echo sounder was a
                 synchronized optical system. The synchronization
                 mechanism between the acoustic and optical systems
                 provided capabilities to have very accurate ground
                 truth recordings for the acoustic data, which were then
                 used as a supervised training data set for the
                 recognition of seaflood vegetation.

                 In this study, results acquired and conclusions made
                 were all based on the comparison against the
                 photographic recordings. The conclusion drawn from this
                 investigation is only as accurate as within the
                 selected habitat types and within very shallow water
                 regions.

                 In order to complete this study, detailed studies of
                 literature and deliberately designed field experiments
                 were carried out. Acoustic data classified with the
                 help of the synchronized optical system were
                 investigated by several methods. Conventional methods
                 such as statistics and multivariate analyses were
                 examined. Conventional methods for the recognition of
                 the collected data gave some useful results but were
                 found to have limited capabilities. When seeking for
                 more robust methods, an alternative approach, Genetic
                 Programming (GP), was tested on the same data set for
                 comparison. Ultimately, the investigation aims to
                 understand potential methods which can be effective in
                 differentiating the acoustic backscatter signals of the
                 habitats observed and subsequently distinguishing
                 between the habitats involved in this study.",
  notes =        "pdf scattered

                 Supervisor Prof. Alexander Gavrilov, Alec Duncan",
}

Genetic Programming entries for Yao-Ting Tseng

Citations