2025_programme: Dynamic region method for improving adaptive quantization during sonar data transmit

Back


  • Day: June 17, Tuesday
      Location / Time: B. ERATO at 17:40-18:00
  • Last minutes changes: -
  • Session: 28. Signal and Image Processing
    Organiser(s): N/A
    Chairperson(s): Gabriele Morra
  • Lecture: Dynamic region method for improving adaptive quantization during sonar data transmit
    Paper ID: 2183
    Author(s): Hyunjae Song, Minho Kim, Youngsub Lim, Youngshin Kim
    Presenter: Hyunjae Song
    Abstract: Sonar system required massive amount of data on long period of time by receiving data repeatedly. Recently, it is necessary that reducing data rate research because high frequency sonar and multi-static sonar. There are many previous study for reduce data rate, sampling reduction, vector quantization. But these studies require to chance exist systems. Adaptive quantization that reduce data rate using exist system has advantage applicable to sonar system. Adaptive quantization divide signal into equal intervals (Uniform region method, URM). It does not consider characteristics of signal. In this paper, I search adaptive quantization bit to fit sonar system and suggest dynamic region method (DRM) that divide region along with characteristics of signal.\n It is calculated differential of signal to divide similar signal into same region. It used data acquired from acoustic sensor that receive every 2 seconds after transmit 0.2 seconds. Short-time Fourier transform (STFT) analysis and percentage error were conducted, and compared with the URM and DRM. \n It is shown that noise increase at STFT analysis in both methods as quantization bit decrease. But there is less noise occurred from 5-bit quantization result. It is also shown that DRM has lower percentage error in every quantization bit, maximum percentage error difference is 22.61 % at 4-bit quantization result.\n We demonstrated that our proposed method is capable of effectively applying adaptive quantization method at sonar system and reduce data rate for improving quantization error. It is necessary that address to compare final image conducted rest of sonar signal processing providing to operator.
      Download the full paper
    This paper is a candidate for the "Prof. John Papadakis award for the best paper presented by a young acoustician(under 40)"
  • Corresponding author: Mr Hyunjae Song
    Affiliation: Underwater Surveillance Technology Research and Development, LIG Nex1 Co, Ltd.,
    Country: Korea, Republic of

Back