【摘要】 将Bivariate模型引入到图像复原中,以Bivariate概率分布函数作为自然图像小波系数向量的先验模型。从图像复原的Bayesian理论出发,提出基于Bivariate概率分布函数非抽取小波域的图像复原算法,并从自适应规整化的角度来分析该算法的有效性。通过对4幅标准测试图像复原实验,并将该算法复原结果与其他3种人们熟知的图像复原算法效果进行对比来证明该算法的有效性。更多还原

【Abstract】 Regards the Bivairate Probability Distribution Function(PDF) as the apriori distribution of wavelet transform coefficients of natural images,and proposes an image restoration algorithm based on Bivariate distributions of un-decimated wavelet domain from the view of Bayesian theory.Discusses this restoration algorithm from the view of local adaptive regularization.At last,verifies this algorithm through performing experiments with four standard testing images and comparing with other some well-known restoration algorithms.更多还原