Adaptive Improved Modified Riesz Mean Filter for Higher Density Salt and Pepper Noise Removal in Grayscale & Color Images

In image processing, median filtering is a widely used technique for removing salt and pepper noise, but it often struggles with preserving image details, especially at high noise levels. In this paper, an enhanced denoising method is presented based on the Modified Riesz Mean (MRM) function, which significantly outperforms median filtering and other state-of-the-art methods, particularly in cases with noise levels between $60 \%$ and $90 \%$. Our approach is effective on both grayscale and color images, and scales efficiently to larger dimensions. The effectiveness of this method lies in its adaptive windowing and reduced computational overhead. It achieves mean SSIM values above 0.88 recorded at high noise density range for both grayscale and color images. These results demonstrate the versatility and the superior performance of the proposed algorithm in denoising and enhancing image quality, making it a reliable choice for image processing tasks.