Hybrid Speckle De-Noising Filters for Ultrasound Images

Kwazikwenkosi Sikhakhane, Suvendi Rimer, Mpho Gololo, Khmaies Ouahada, Adnan Abu-Mahfouz

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Ultrasound imaging uses reflected echo signals from high-frequency sound waves transmitted into a body with high water content to create images of the internal soft tissue, bone, and blood flow. Different kinds of noise are introduced during the signal acquisition and processing stages. Ultrasound images are mostly affected by speckle noise. High-quality images can improve the observational accuracy of diagnostic exams. Speckle noise degrades an image's quality, making it difficult to recognize, analyze, and measure. To reduce speckle noise in ultrasound images and improve diagnostic capability, an investigation and comparison study of the different image filters currently available was done. Thereafter a hybrid filter combination of speckle, gaussian, median and average filter was used to remove speckle noise in ultrasound images. The proposed hybrid filters outperformed other filters in terms of PSNR, SNR, SSIM and MSE.

Original languageEnglish
Title of host publicationProceedings of the 16th IEEE AFRICON, AFRICON 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798350336214
DOIs
Publication statusPublished - 2023
Event16th IEEE AFRICON, AFRICON 2023 - Nairobi, Kenya
Duration: 20 Sept 202322 Sept 2023

Publication series

NameIEEE AFRICON Conference
ISSN (Print)2153-0025
ISSN (Electronic)2153-0033

Conference

Conference16th IEEE AFRICON, AFRICON 2023
Country/TerritoryKenya
CityNairobi
Period20/09/2322/09/23

Keywords

  • digital filters
  • Gaussian
  • image processing
  • Median
  • speckle noise
  • Specklefilt
  • Wiener

ASJC Scopus subject areas

  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Hybrid Speckle De-Noising Filters for Ultrasound Images'. Together they form a unique fingerprint.

Cite this