Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Joint demosaicing and denoising.

Keigo Hirakawa1, Thomas W Parks

  • 1Cornell University, Ithaca, NY 14853, USA. hirakawa@stat.harvard.edu

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|August 12, 2006
PubMed
Summary

This study introduces a unified method for digital camera image processing, combining demosaicing and denoising into one step. This approach significantly enhances image quality by reducing sensor noise while accurately reconstructing color information.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

BayesToF: multiresolution denoising of indirect time-of-flight distance maps.

Optics express·2026
Same author

Stokes Simplex Modeling for Polarization Image Denoising.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2025
Same author

Centroiding Point-Objects With Event Cameras.

IEEE transactions on pattern analysis and machine intelligence·2025
Same author

EBSnoR: Event-Based Snow Removal by Optimal Dwell Time Thresholding.

IEEE transactions on pattern analysis and machine intelligence·2025
Same author

Event-Based Visual Microphone Based on Specular Reflections Off Angularly Deformed Surfaces.

IEEE transactions on pattern analysis and machine intelligence·2025
Same author

LMP-GAN: Out-of-Distribution Detection for Non-Control Data Malware Attacks.

IEEE transactions on pattern analysis and machine intelligence·2025

Area of Science:

  • Digital image processing
  • Computational imaging
  • Signal processing

Background:

  • Digital camera images suffer from sensor noise, degrading output quality.
  • Demosaicing and denoising are typically separate, sequential image processing steps.
  • Existing methods often fail to optimally address noise during demosaicing.

Purpose of the Study:

  • To develop a novel, unified technique for simultaneous demosaicing and denoising.
  • To leverage the inherent similarities between demosaicing and denoising operations.
  • To improve overall digital image quality by integrating these processes.

Main Methods:

  • Designed a filter for optimal pixel estimation from noisy sensor data.
  • Adapted filter coefficients for demosaicing (color filter array interpolation) under noisy conditions.

Related Experiment Videos

  • Integrated a total least squares denoising method to demonstrate the unified approach.

    • Main Results:

    • The proposed method effectively suppresses sensor noise (e.g., CMOS/CCD noise models).
    • Simultaneous processing achieved superior interpolation of missing pixel color components.
    • Experimental validation on synthetic and real CMOS camera data confirmed significant image quality improvements.

    Conclusions:

    • Combining demosaicing and denoising into a single operation is highly effective.
    • The unified approach outperforms independent processing of these tasks.
    • This technique offers a systematic way to integrate various denoising algorithms with demosaicing.