<p>Restoration of images degraded with mixed noise is an important problem in digital image processing. In this paper, we address the problem of mixed Gaussian noise and impulse noise pruning for color images, and review the manifold algorithms stacked in the literature. Edge preservation is a crucial challenge while denoising an image. Bilateral filters and their extensions are comprehensively used in noise reduction while preserving edges. However, several impediments are hindering their further development. We propose a novel filtering scheme based on the concept of outlier mitigation between pixels using local image statistics. The introduced novel weighting function is a new weight function for bilateral filtering methods, combining the concept of winsorized mean to clean outliers and conventional topographic closeness. The performance of the proposed scheme is compared with widely used denoising schemes. The experimental outcomes, validated on different types of standard images, prove that the execution of the proposed scheme is upright and more acceptable than the competent schemes in terms of standard visual quality parameters, such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), mean absolute error (MAE), and mean square error (MSE).</p>