Noise Estimation by Utilizing Mean Deviation of Smooth Region in Noisy Image

In many practical cases of image processing, only a noisy image is available. Many image denoising methods usually require the exact value of the noise distribution as an essential filter parameter. However, to estimate the noise solely from the information of the noisy image is a difficult task. A simple but accurate noise estimator would significantly benefit many image denoising methods. In this paper, we present a method to estimate additive noise by utilizing the mean deviation of a smooth region selected from a noisy image. The noise distribution is estimated by computing the average mean deviation of all non-overlapping blocks in the smooth region. Simulation results demonstrate that our method achieves accurate noise estimation regardless of the image characteristics over a range of noise levels. The restoration performance of a denoising technique based on our noise estimation method resembles that achieved in the ideal condition.