基于状态连续变化的Hopfield神经网络的图像复原  

Image Restoration Using a Modified Hopfield Neural Network of Continuous State Change

作　　者：韩玉兵[1] 吴乐南[1] 

机构地区：[1]东南大学无线电工程系,南京210096

出　　处：《信号处理》2004年第5期431-435,共5页Journal of Signal Processing

摘　　要：针对图像复原提出了神经元状态连续变化的Hopfield神经网络模型,详细讨论了两种连续函数串行、全并 行复原算法的收敛性和参数选择,仿真实验表明,该模型能够精确达到能量极小点,并对复原图像的信噪比有一定的提高。A modified Hopfield neural network model based on continuous state change is proposed to restore a degraded image. In this paper, we discuss the serial and the parallel algorithm of two continuous functions respectively, and thoroughly study the convergence and the choice of parameters. Experimental results demonstrate that this model can obtain the minimum of the energy and the SNR of the restored image has some improvement.

关 键 词：图像复原 HOPFIELD神经网络模型 并行 串行 算法 仿真实验 状态 极小点 连续函数 收敛性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391[自动化与计算机技术—控制科学与工程]