为了保护图像边缘、细节等信息,建立了l_1范数约束的非局部均值正则模型.首先通过实验分析了非局部均值去噪算法余项的概率密度函数具有较强的拖尾性质,符合Laplace分布的特点.基于此,使用l_1范数约束的非局部均值去噪算法余项作为新的正则项,提出了新的正则模型.然后利用Bregman算子分裂算法求解得到了相应的优化算法,并且新算法可看成Plug-and-Play Priors算法的推广.实验结果表明新模型在去除模糊,保护图像边缘、细节等信息方面性能都优于l_2范数约束的非局部均值正则模型及Plug-and-Play Priors模型. In order to protect the edge and detail of the image, a nonlocal mean regular model with l_1 norm constraint was established.Firstly, the probability density function of the remainder of the non-local mean noise denoising algorithm was analyzed by experiments, Based on this, a new regular model is proposed using the remainder of nonlocal norm denoising algorithm with l-1 norm constraint as a new regular item.And then the corresponding optimization algorithm is obtained by using Bregman operator splitting algorithm, and new The algorithm can be regarded as popularization of Plug-and-Play Priors algorithm.Experimental results show that the performance of the new model is better than l_2 norm-constrained nonlocal mean regular model and the plug-and- Play Priors model.