Regeneration Filter: Enhancing Mosaic Algorithm for Near Salt & Pepper Noise Reduction.

Area of Science:

• Image processing
• Computer vision
• Digital signal processing

Background:

• Digital images are often corrupted by noise, such as near Salt-and-Pepper (nS&P) noise, which degrades visual quality and hinders analysis.
• Conventional noise reduction filters may blur image details or fail to effectively remove high-density noise.
• Existing methods often rely on median or other complex filtering techniques, necessitating alternative approaches.

Purpose of the Study:

• To develop and present a novel Regeneration filter for efficient and selective removal of nS&P noise from digital images.
• To demonstrate the filter's capability in preserving structural image details during the noise reduction process.
• To provide a robust method for image denoising that outperforms conventional techniques, especially under high noise conditions.

Main Methods:

• A new Regeneration filter is proposed, focusing on restoring noise-affected pixels via localized contextual analysis.
• The filter utilizes an iterative processing approach, designed to avoid image quality degradation with increased iterations.
• Performance evaluation is conducted using standard image quality assessment metrics and experimental comparisons.

Main Results:

• The Regeneration filter effectively reduces nS&P noise while preserving essential image structures.
• The filter demonstrates robustness and consistent high-quality results even with noise densities up to 97%.
• Iterative processing does not negatively impact image quality, even at high noise levels.

Conclusions:

• The proposed Regeneration filter offers a superior method for nS&P noise reduction in digital images.
• The filter's ability to preserve details and handle extreme noise levels makes it a valuable tool in image processing.
• The availability of code and data in a public repository ensures transparency and facilitates further research.