Characterising the statistics of wavelet coefficients is a critical issue i
n image compression and denoising. Many powerful approaches have been inves
tigated, but accurate modelling suffers From high computation complexity. I
n this work an efficient adaptive algorithm to capture the dependency of bo
th inner and inter scale wavelet coefficients is proposed. Experimental res
ults show that compared with the algorithm in [3], in the case of higher no
ise variance, greater PSNR performance gain may be obtained.