Fast and reliable noise estimation algorithm based on statistical hypothesis tests
Ping Jiang, Jianzhou Zhang · 2012
Image noise estimation is a very important topic in digital image processing. This paper presents a fast and reliable noise estimation algorithm for additive white Gaussian noise (WGN). The proposed algorithm provides a way to measure the degree of image feature based on statistical hypothesis tests (SHT). Firstly, the proposed algorithm distinguishes homogeneous blocks and non-homogeneous blocks by the degree of image feature, and then sets the minimal variance of these homogeneous blocks as a reference variance. Secondly, the proposed algorithm finds more homogeneous blocks whose variances are similar to the reference variance and which are not non-homogeneous blocks. Lastly, the noise variance is estimated from these homogeneous blocks by a weighted averaging process according to the degree of image feature. Experiments show that the proposed algorithm performs well and reliably for different types of images over a large range of noise levels.