• DocumentCode
    1098781
  • Title

    A physical model-based analysis of heterogeneous environments using sonar-ENDURA method

  • Author

    Bozma, Ömür ; Kuc, Roman

  • Author_Institution
    Dept. of Electr. Eng., Yale Univ., New Haven, CT, USA
  • Volume
    16
  • Issue
    5
  • fYear
    1994
  • fDate
    5/1/1994 12:00:00 AM
  • Firstpage
    497
  • Lastpage
    506
  • Abstract
    A physical model-based analysis of unstructured environments is presented using sonar as a sensing device. Previous methods have relied only on time-of-flight (TOF) methods and have examined only homogenous environments consisting of either smooth or rough surfaces. In this paper, a forward model for the reflection from a class of surfaces with varying degrees of roughness is presented based on the Kirchhoff approximation method. This model integrates different types of environments into a single analytical framework. The echo intensity is parametrized in terms of its energy content and duration, which are functions of the surface roughness, distance, and orientation. The echo-energy and echo-duration maps are introduced to display these parameters. A systematic and robust procedure (ENDURA method) is presented to analyze the reflections and to differentiate and localize the reflecting surfaces. The methodology is verified with experimental results obtained in our laboratory. The results indicate a significant improvement over conventional TOF systems
  • Keywords
    acoustic signal processing; sonar; ENDURA method; Kirchhoff approximation method; distance; echo intensity parametrization; echo-duration maps; echo-energy maps; energy content; energy duration; heterogeneous environments; orientation; physical model-based analysis; sonar; surface roughness; time-of-flight methods; unstructured environments; Acoustic scattering; Displays; Intelligent sensors; Kirchhoff´s Law; Laboratories; Reflection; Robustness; Rough surfaces; Sonar; Surface roughness;
  • fLanguage
    English
  • Journal_Title
    Pattern Analysis and Machine Intelligence, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0162-8828
  • Type

    jour

  • DOI
    10.1109/34.291448
  • Filename
    291448