Impulse Noise Removal from Medical Images by Two Stage Quaternion Vector Median Filter

Area of Science:

• Medical Imaging
• Image Processing
• Computational Mathematics

Background:

• Impulse noise significantly degrades medical image quality, hindering accurate diagnosis.
• Existing noise reduction filters often struggle to balance noise suppression with detail preservation.

Purpose of the Study:

• To develop and evaluate a novel two-stage filter for effective impulse noise removal in medical images.
• To leverage quaternion representation for enhanced processing of both grayscale and color medical images.

Main Methods:

• A two-stage filtering approach utilizing quaternion representation for pixel differences.
• Pixel ranking based on aggregated differences within a filtering window.
• Impulse noise detection using rank and threshold criteria.
• Noisy pixel replacement via a quaternion-based vector median filter.

Main Results:

• The proposed filter demonstrated superior performance in suppressing impulse noise compared to established methods.
• The filter effectively preserved fine details and structural information in medical images.
• Quaternion processing maintained the integrity of color image channels, treating them as a unified entity.

Conclusions:

• The developed quaternion-based filter offers a robust solution for impulse noise removal in medical imaging.
• This method provides an effective means to enhance the quality of medical images for diagnostic purposes.
• The approach shows promise for application in various medical imaging modalities.