基于自适应字典学习的乘性噪声去除模型

摘要

In order to remove the noise of image,a new image multiplicative noise removal model based on sparse representation and adaptive dictionary is proposed.PCA dictionary and iteration shrinkage algorithm are used to update the sparse code.Newton-iteration method is used to obtain the restored image of log-domain.By an exponential function and error correction,the denoising image is obtained in the real domain.Experimental results demonstrate that compared with several existing noise suppression algorithms,this model can hold important information of the image better while effectively removing multiplicative noise.%为了更好地去除图像中的噪声,提出一种基于稀疏表示的自适应字典学习的图像乘性噪声抑制模型.利用PCA字典和迭代收缩算法更新稀疏编码,用牛顿迭代法获得对数域中的恢复图像,再利用指数函数以及误差校正将得到的结果转到实数域中.实验结果表明,与已有的4种抑噪算法相比,该模型在有效去除乘性噪声的同时,能够更好地保持原始图像的重要信息.