关键词: 脉冲耦合神经网络, 高斯噪声滤波, 自适应性

Abstract: A new approach of image filtering is presented in this paper concerning Gauss noise with high deviation. This approach firstly locates large noise pixels and only filters these pixels using an analogous average filter, then smoothes small noise pixels through improved Pulse-coupled neural networks. The proposed method works well, and has strong adaptability as well as protecting image edge compared with other filtering methods. The experiment results and comparisons show that this method is feasible and effective.

Key words: Pulse-coupled neural networks（PCNN）, Gauss noise filter, Adaptive