Background/Objectives: Noise reduction is an essential step in visual improvement task and it plays a major role in subsequent processing tasks such as image analysis. Thus, the main purpose of image denoising especially in natural image is to suppress the artifacts which destroy an image and suppress the additive Gaussian noise without losing the delicate texture of the latent image. Methods/Statistical Analysis: Basically, the major the noise which attained in the processes of acquisition and diffusion of several digital images is presumed to be AWGN. Accordingly, a novel framework is proposed based on lifting scheme approach in wavelet domain by utilizing hidden Markov model. In this context, statistical model which is represents by HMM plays a prominent role. It is due to the ability to capture the dependencies and the correlations between the coefficients in different decomposition levels. Findings: The extensive experiments prove the efficiency of SGWs-HMM in several standard images when compared with best state-of-the-art noise removal methods. SGWs-HMM shows high competitive performance in terms of visual quality and subjective assessments with the widely used denoising techniques. Application/Improvements: This approach shows comparable improvement in subjective and objective assessments. Regarding to the computation complexity, the proposed algorithm is more expensive than rest of algorithms under investigation due to the statistical structure of the proposed method.