Abstract: We analysed  the parametric accuracy of the generalized noise model estimated based on the maximum likelihood equation by using Fisher information and the associated asymptotic normality. The theoretical analysis results show that for standard pixel images, the parameter error of the additive noise estimated by using the maximum likelihood equation is larger than the error of signal-dependent noise parameter, while for the normalized images, the accuracy of the parameters is exactly the opposite. The experiments prove the correctness  of the theoretical analysis.

Key words: parametric accuracy, maximum likelihood equation, Fisher information, asymptotic normality