A modified form of different applied median filter for removal of salt & pepper noise

• Published in: Multimedia tools and applications Vol. 82; no. 5; pp. 7479 - 7490
• Main Authors: Aslam, Numan, Ehsan, Muhammad Khurram, Rehman, Zia Ul, Hanif, Muhammad, Mustafa, Ghulam
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.02.2023; Springer Nature B.V
• Subjects: Algorithms; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Multimedia; Multimedia Information Systems; Pixels; Signal to noise ratio; Special Purpose and Application-Based Systems; Impulse noise; Random noise; Median filter; Salt & pepper noise
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-022-13289-x

An algorithm is presented for removal of Salt & pepper noise. Proposed algorithm uses two phased approach where noise is first detected and then removed using window sizes extending up to 7 × 7. Window size changes depending upon noise density. In any processing window, we have some pre-processed pixels and some unprocessed pixels. If a pixel under consideration is a noise pixel then we consider all processed and unprocessed pixels for noise pixel replacement, if we find one or more noise-free pixels in window then we replace median of these pixels with corrupted pixel. If we do not find any original pixel in window then we increase window size and repeat retrospective process. The proposed algorithm shows improved results as compared to existing methods. Proposed algorithm is tested and compared for Structural Similarity (SSIM) and Peak-Signal-to-Noise Ratio (PSNR) with existing methods.