Frequency Domain Characteristics Analysis of Non-Coupled PCNN

doi:10.19678/j.issn.1000-3428.0061296

Abstract

Abstract: The Pulse Coupled Neural Network (PCNN) model is characterized by pulse modulation and coupling connection, which is widely used in the field of digital image processing.However, the existing research on the PCNN model focuses on analyzing the relationship between the parameters and model characteristics from the perspective of time domain or the information contained in the image itself, which cannot fully explain the influence of the parameters on the model characteristics.Starting from the iterative equation of the PCNN model, this study analysis the non-coupled PCNN model in the frequency domain using the discrete system frequency domain analysis method.The results revealed that the dynamic threshold subsystem of the non-coupled PCNN model has low-pass characteristics, and the selection range of model parameters a_E is determined.Meanwhile, through the derivation of the pulse firing time formula, the selection range of the parameters v_E is obtained.The Fourier transform method is used to analyze the pulse firing frequency characteristics and dynamic threshold attenuation frequency characteristics of a single neuron, and to explain the influence of the network parameters of non-coupled PCNN model on the frequency domain characteristics of the network.The simulation results verify the correctness of the theoretical analysis' conclusion.The theoretical analysis clarifies the relationship between non-coupled PCNN model parameters and network characteristics from the perspective of frequency domain and provides a new method for mining non-coupled PCNN model characteristics.

Key words: Pulse Coupled Neural Network(PCNN), dynamic threshold subsystem, Fourier transform, pulse firing frequency, dynamic threshold attenuation frequency

摘要： 脉冲耦合神经网络（PCNN）模型具有脉冲调制和耦合连接特性，广泛应用于数字图像处理领域。然而现有PCNN模型的研究都是从时域或图像本身包含信息角度分析参数与模型特性之间的关系，无法全面解释参数对模型特性的影响。从PCNN模型的迭代方程出发，利用离散系统频域分析方法从频域角度对无耦合PCNN模型进行分析，揭示无耦合PCNN模型的动态门限子系统具有低通特性，并确定网络参数a_E的选取范围，同时通过对脉冲发放时刻公式进行推导，得到参数v_E的选取范围。采用傅里叶变换方法分析单个神经元的脉冲发放频率特性和动态门限衰减频率特性，解释无耦合PCNN模型的参数对频域特性的影响。仿真实验结果验证了该理论分析结论的正确性，从频域角度理解无耦合PCNN模型的参数与模型特性之间的关系，为挖掘PCNN模型特性提供一种新的方法。

关键词: 脉冲耦合神经网络, 动态门限子系统, 傅里叶变换, 脉冲发放频率, 动态门限衰减频率

CLC Number:

TP183
TP391

DENG Xiangyu, Lü Yahui, CHEN Yan. Frequency Domain Characteristics Analysis of Non-Coupled PCNN[J]. Computer Engineering, 2022, 48(6): 213-221.

邓翔宇, 吕亚辉, 陈岩. 无耦合PCNN频域特性分析[J]. 计算机工程, 2022, 48(6): 213-221.

/ / Recommend / Download Citations

URL: http://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0061296

http://www.ecice06.com/EN/Y2022/V48/I6/213

Figures/Tables 9

References

[1] ECKHORN R, REITBOECK H J, ARNDT M, et al.Feature linking via synchronization among distributed assemblies:simulations of results from cat visual cortex[J].Neural Computation, 1990, 2(3):293-307.
[2] JOHNSON J L, PADGETT M L.PCNN models and applications[J].IEEE Transactions on Neural Networks, 1999, 10(3):480-498.
[3] LIAN J, YANG Z, LIU J Z, et al.An overview of image segmentation based on pulse-coupled neural network[J].Archives of Computational Methods in Engineering, 2021, 28(2):387-403.
[4] GUO Y N, YANG Z, MA Y D, et al.Saliency motivated improved simplified PCNN model for object segmentation[J].Neurocomputing, 2018, 275(C):2179-2190.
[5] TAN W, XIANG P, ZHANG J J, et al.Remote sensing image fusion via boundary measured dual-channel PCNN in multi-scale morphological gradient domain[J].IEEE Access, 2020, 8:42540-42549.
[6] LI L L, MA H B.Pulse coupled neural network-based multimodal medical image fusion via guided filtering and WSEML in NSCT domain[J].Entropy, 2021, 23(5):591.
[7] MOHAMMED M M, BADR A, ABDELHALIM M B.Image classification and retrieval using optimized Pulse-Coupled Neural Network[J].Expert Systems With Applications, 2015, 42(11):4927-4936.
[8] YONEKAWA M, KUROKAWA H.The content-based image retrieval using the pulse coupled neural network[C]//Proceedings of International Joint Conference on Neural Networks.Washington D.C., USA:IEEE Press, 2012:1-8.
[9] DENG X Y, MA Y D, DONG M.A new adaptive filtering method for removing salt and pepper noise based on multilayered PCNN[J].Pattern Recognition Letters, 2016, 79:8-17.
[10] SHEN C, WANG D, TANG S M, et al.Hybrid image noise reduction algorithm based on genetic ant colony and PCNN[J].The Visual Computer, 2017, 33(11):1373-1384.
[11] DENG X Y, MA Y.PCNN model analysis and its automatic parameters determination in image segmentation and edge detection[J].Chinese Journal of Electronics, 2014, 23(1):97-103.
[12] ZHOU L, SUN Y, ZHENG J G.Automated color image edge detection using improved PCNN model[J].WSEAS Transactions on Computers, 2008, 7(4/6):186-191.
[13] DENG X Y, YANG Y H, QIN W J.An improved non-coupled PCNN model for image segmentation[C]//Proceedings of the 2nd International Conference on Communication, Network and Artificial Intelligence.Guangzhou, China:[s.n.], 2019:1-10.
[14] HELMY A K, EL-TAWEEL G S.Image segmentation scheme based on SOM-PCNN in frequency domain[J].Applied Soft Computing, 2016, 40:405-415.
[15] CHINMAYA P, AYAN S, KUMAR M N.Fractal dimension based parameter adaptive dual channel PCNN for multi-focus image fusion[J].Optics and Lasers in Engineering, 2020, 133:10614.
[16] 闫利, 向天烛.NSCT域内结合边缘特征和自适应PCNN的红外与可见光图像融合[J].电子学报, 2016, 44(4):761-766. YAN L, XIANG T Z.Fusion of infrared and visible images based on edge feature and adaptive PCNN in NSCT domain[J].Acta Electronica Sinica, 2016, 44(4):761-766.(in Chinese)
[17] CHEN Y L, SUNG-KEE P, MA Y D, et al.A new automatic parameter setting method of a simplified PCNN for image segmentation[J].IEEE Transactions on Neural Networks, 2011, 22(6):880-92.
[18] MA Y D, LIU Q, QIAN Z B.Automated image segmentation using improved PCNN model based on cross-entropy[C]//Proceedings of International Symposium on Intelligent Multimedia, Video and Speech Processing.Washington D.C., USA:IEEE Press, 2004:743-746.
[19] 姚畅, 陈后金, 李居朋.改进型脉冲耦合神经网络在图像处理中的动态行为分析[J].自动化学报, 2008, 34(10):1291-1297. YAO C, CHEN H J, LI J P.Analysis of dynamic behaviors of improved pulse coupled neural network in image processing[J].Acta Automatica Sinica, 2008, 34(10):1291-1297.(in Chinese)
[20] 郝敬帅, 高媛, 程橙, 等.对脉冲耦合神经网络中被动神经元的脉冲周期分析[J].计算机测量与控制, 2019, 27(6):256-262. HAO J S, GAO Y, CHENG C, et al.Pulse cycle analysis of passive neurons in the pulse coupled neural network[J].Computer Measurement & Control, 2019, 27(6):256-262.(in Chinese)
[21] DENG X Y, YAN C M, MA Y D.PCNN mechanism and its parameter settings[J].IEEE Transactions on Neural Networks and Learning Systems, 2020, 31(2):488-501.
[22] MA Y D, SHI F, LI L.Gaussian noise filter based on PCNN[C]//Proceedings of International Conference on Neural Networks and Signal Processing.Washington D.C., USA:IEEE Press, 2003:149-151.
[23] 邓翔宇, 马义德.PCNN参数自适应设定及其模型的改进[J].电子学报, 2012, 40(5):955-964. DENG X Y, MA Y D.PCNN model automatic parameters determination and its modified model[J].Acta Electronica Sinica, 2012, 40(5):955-964.(in Chinese)

Please choose a citation manager

Content to export