Adaptive conductance function based improved diffusion filtering and bi-dimensional empirical mode decomposition based image denoising

Abstract

This paper presents a new method for image denoising based on a two-dimensional empirical mode decomposition algorithm and semi-adaptive diffusion coefficient in anisotropic diffusion filter. The proposed model uses a local difference value method to compare and replace some pixels of the noisy image with a pre-processed image that has been passed through a Gaussian filter. A bi-dimensional empirical mode decomposition algorithm is then employed to decompose the noise-contaminated image into its intrinsic mode functions in which high-frequency and low-frequency noise components are removed by applying a diffusion filter. The filter has a semi-adaptive threshold in the diffusion coefficient with parameters like connectivity, conductance function, number of iterations, and gradient threshold. The semi-adaptive threshold for each diffusion is implemented by introducing gradient values in the threshold of the corrupted image. The image is then reconstructed from these denoised intrinsic mode functions. The performance of the proposed method is assessed in terms of peak signal-to-noise ratio, mean square error, and structural similarity index and is compared with the existing methodologies. The results obtained from experimentation indicate that the proposed method is efficient in both feature retention and noise suppression. © 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.