An adaptive edge model in the wavelet domain for wavelet image coding

Ateş, Hasan Fehmi and Orchard, Michael T. (2005) An adaptive edge model in the wavelet domain for wavelet image coding. Signal Processing: Image Communication, 20 (2). pp. 169-185. ISSN 0923-5965

[img]PDF - Registered users only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader

Official URL: http://dx.doi.org/10.1016/j.image.2004.11.001


State-of-art wavelet coders owe their performance to smart ideas for exploiting inter and intra-band dependencies of wavelet coefficients. We claim that developing more efficient coders requires us to look at the main source of these dependencies; i.e., highly localized information around edges. This paper investigates the structural relationships among wavelet coefficients based on an idealized view of edge behavior, and proposes a simple edge model that explains the roots of existing dependencies. We describe how the model is used to approximate and estimate the significant wavelet coefficients. Simulations support its relevance for understanding and analyzing edge information. Specifically, model-based estimation within the space-frequency quantization (SFQ) framework increases the peak signal-to-noise ratio (PSNR) by up to 0.3dB over the original SFQ coding algorithm. Despite being simple, the model provides valuable insights into the problem of edge-based adaptive modeling of value and location information in the wavelet domain.

Item Type:Article
Uncontrolled Keywords:wavelet subband; edge contour; adaptive model; image coding; interpolation
Subjects:T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101-6720 Telecommunication
ID Code:13987
Deposited By:Hasan Fehmi Ateş
Deposited On:31 May 2010 17:09
Last Modified:25 Jul 2019 09:41

Repository Staff Only: item control page