Vector processing of wavelet coefficients for robust image denoising

M. E. Zervakis, V. Sundararajan, K. K. Parhi

Research output: Contribution to journalArticlepeer-review

11 Scopus citations


This paper presents a wavelet-domain robust denoising algorithm, which efficiently removes both Gaussian as well as Gaussian mixed with impulse noise. Several wavelet domain operators that help in the robust denoising process are developed. Robustness of the various operators is established from a practical viewpoint with the help of simulation. Quantitative and qualitative performance comparisons of the proposed operators and existing wavelet and image domain-based denoising methods are presented. The superiority of the new scheme is established by simulation results over a variety of images. A novel approach to implement new operators that simplify hardware implementation is presented. Implementation efficiency is established mainly from an area/power and partly from a speed perspective. Simulations are carried out under Gaussian and mixed (Gaussian + Impulse) noise contamination to demonstrate the robustness of the proposed approach.

Original languageEnglish (US)
Pages (from-to)435-450
Number of pages16
JournalImage and Vision Computing
Issue number7
StatePublished - May 1 2001


  • Gaussian
  • Robust image denoising
  • Vector processing


Dive into the research topics of 'Vector processing of wavelet coefficients for robust image denoising'. Together they form a unique fingerprint.

Cite this