%0 Conference Paper
%A Kamel, N.
%A Kafa, S.
%D 2010
%F scholars:943
%K Additive white noise; AR models; Auto-correlation value; Auto-regressive; Autoregressive modeling; Cross-correlation function; Linear prediction; Predictor coefficients; Second order statistics; Signal to noise ratio estimation; Single images; SNR estimation; SNR values, Computer simulation; Signal to noise ratio; White noise, Estimation
%R 10.1109/ICIAS.2010.5716130
%T Image SNR estimation using the autoregressive modeling
%U https://khub.utp.edu.my/scholars/943/
%X A number of techniques have been proposed during the last two decades for Signal-to-Noise Ratio (SNR) estimation in images. The majority of these techniques are based on the cross-correlation function of two images of the same area. However, the need for two images to estimate SNR value confines these techniques to non-stored images and thus limits their applications. In this paper the second order statistics of image corrupted by additive white noise are modeled by Autoregressive-model and the relationship between AR model and linear prediction is utilized in estimating the predictor coefficients. The predictor is then used to estimate the zero-offset autocorrelation value and accordingly obtain the power of the noise-free image. Unlike others, the proposed technique is based on single image and offers the required accuracy and robustness in estimating the SNR values.
%Z cited By 4; Conference of 2010 International Conference on Intelligent and Advanced Systems, ICIAS 2010 ; Conference Date: 15 June 2010 Through 17 June 2010; Conference Code:84196