본문 바로가기
  • Home

Quantization Noise Reduction Algorithm Using Wavelet Coefficients Characteristic and Signal Adaptive Filter

  • Journal of Knowledge Information Technology and Systems
  • Abbr : JKITS
  • 2012, 7(1), pp.129-137
  • Publisher : Korea Knowledge Information Technology Society
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Published : February 29, 2012

Kyung Nam Park 1

1나사렛대학교

Accredited

ABSTRACT

In this paper, we propose a postprocessing algorithm, which reduces the quantization noises in block-coded images by using wavelet coefficients and signal adaptive filter. The proposed algorithm performs block classification as smooth region and complex region using the magnitude and phase of wavelet transform coefficients. In smooth region on the block boundary filtering is perform to remove grid noise. Complex region is classified into edge and staircase noise by using the variation ratio of wavelet coefficient and remove staircase noise only. We removes ringing artifacts using signal adaptive filter in edge blocks of complex region. Simulation results show that the proposed algorithm gives better results than the conventional algorithm from both a subjective and an objective viewpoint.

Citation status

* References for papers published after 2023 are currently being built.