A New Robust Scheme for Salt and Pepper Noise Filtering Using Fuzzy Cellular Automata

<title>Abstract</title> <p>This paper presents a novel and robust scheme for removing salt and pepper noise from digital images using a combination of three newly developed algorithms. The proposed method operates in two main phases: noise detection and noise removal. In the first phase, salt and pepper noise pixels are identified through a two-step process. Initially, a new algorithm detects noisy pixels based on the maximum, minimum, and average values of pixels within the local neighborhood of the target pixel. To enhance detection accuracy, a secondary algorithm re-evaluates the non-identified pixels by computing the Hamming distance of neighboring pixel values using cellular automata principles. In the second phase, the corrupted pixels are restored via a novel algorithm that leverages four distinct six-pixel patterns in the Moore neighborhood, utilizing Jaccard similarity measures integrated with fuzzy cellular automata. The performance of the proposed method is quantitatively assessed using Normalized Cross-Correlation (NCC), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Image Enhancement Factor (IEF). Experimental results demonstrate that the proposed approach is simple, robust, and highly parallelizable, effectively preserving fine image details while outperforming existing techniques.</p>