基于耦合神经网络反应扩散模型的图像增强方法研究  

Research on image enhancement method based on coupled neural network reaction diffusion model

作　　者：周嵩松 李平[1] 赵文博 王行建[1] Zhou Songsong;Li Ping;Zhao Wenbo;Wang Xingjian(College of Computer and Control Engineering,Northeast Forestry University,Harbin 150040,China)

机构地区：[1]东北林业大学计算机与控制工程学院,哈尔滨150040

出　　处：《现代计算机》2024年第10期55-59,64,共6页Modern Computer

基　　金：东北林业大学省级大学生创新训练项目(S202310225009);黑龙江省自然科学基金联合项目(LH2020C048)。

摘　　要：随着人工智能技术的迅速发展,图像处理已经广泛应用于交通、医疗、农林、航空航天等领域,并且对图像质量要求也越来越高。针对图像增强问题,提出了一种利用神经网络模型,将FitzHugh-Nagumo反应扩散模型转变为微分动力系统的方法,来改进耦合神经网络,引入自适应的阈值,得到耦合神经元的非线性动力学模型,利用此模型可使图像对比度得到提升,起到增强图像的效果。该模型的适用面更广,图像对比度拉伸效果更好,视觉增强效果更为明显。With the rapid development of artificial intelligence technology,image processing has been widely used in the fields of transport,medical,agriculture and forestry,aerospace,etc.,and the requirements for image quality are getting higher and higher.For the image enhancement problem,the neural network model is used,and it is proposed to transform the FitzHugh⁃Nagumo reaction⁃diffusion model into a differential dynamics system,so as to improve the coupled neural network,introduce the adaptive threshold,and obtain the nonlinear dynamics model of the coupled neuron,using which the image contrast can be made to improve,and play a role in the enhancement of the image.The model is more widely applicable,the image contrast stretching effect is better,and the visual enhancement effect is more obvious.

关 键 词：耦合神经网络 反应扩散模型 FitzHugh-Nagumo 图像增强 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]