基于小波域统计模型的纸浆纤维图像去噪研究

Image Denoising of Pulp Fiber Based on Statistical Modeling in Wavelet Domain

在小波多尺度分析基础上,提出一种新的图像小波系数的自适应统计算法,并应用于纸浆纤维图像的去噪研究。将图像小波系数视为服从广义高斯分布(GGD)的随机变量模型,在小波软阈值去噪的基础上引入空间自适应阈值方法;将均值滤波算法应用于小波系数方差的边缘估计中,结合最大后验概率准则(M AP)进行参数估计以恢复噪音小波图像。该算法用于纸浆纤维图像的去噪,效果理想,同其它的图像去噪算法相比,它具有较高的峰值信噪比(PSNR)。

An adaptive statistical model for wavelet image coefficients is presented based on the multiscale analysis of wavelet transform,and applied for image denoising of pulp fiber, Each wavelet coefficient is modeled as a random variable of a generalized Gaussian distribution （GGD）. A spatially adaptive wavelet thresholding method based on the soft thresholding method is introduced,the marginal prior distribution of wavelet coefficient variances is estimated by using average filtering algorithm, with maximum posteriori probability rule （MAP）, then they are applied to restore the noisy wavelet image. The algorithm is applied to denoise the noisy pulp fiber image. Simulation results show that higher peak-signal to noise ratio can be obtained as compared to other recent image denoising methods.