Suppression of Gaussian noise using Bayesian classifier in an image

In this paper, a spatially adaptive denoising algorithm is proposed which provides satisfactory results when the image is corrupted with the additive white Gaussian noise. Generally, suppression of Gaussian noise poses a trade-off problem between suppressing the noise and preserving the detailed information of the image. Detection of noise and suppression of noise are the two stages used in the proposed algorithm. Noise detection in the image is modeled as a pattern classification problem. Bayes classifier has been utilized to classify the pixels as noisy or non noisy. For effective noise suppression, a Gaussian filter is used in the proposed method which preserves detailed information as compared to PWMAD, SAWM and SADA methods. Bayesian classifier, computational cost, error detection, over-smoothness, and smoothing degree of reconstructed image are the parameters taken into account to effectively suppress the noise components in the proposed method.