A Statistical Analysis on Differential Signals for Noise Level Estimation
Weihai Li, Bo Wang · 2007
A new noise level estimating method based on a statistical analysis is proposed here. In this method, the noise characteristics are obtained by matching theoretical distribution curve to differential signal histogram. As an example, the problem is addressed to additive white Gaussian noise in digital images, and the noise variance is estimated. The efficiency of this method is verified by the accurate experiment results.