真实场景中基于卷积神经网络的去噪算法  

Denoising Algorithm Based on Convolutional Neural Network in Real Scene

作　　者：谢舒闽 邓磊[1] 谢中梅[1] 李海平 XIE Shumin;DENG Lei;XIE Zhong-mei;LI Hai-ping(Jiangxi College of Applied Technology,Ganzhou 341000,Jiangxi)

机构地区：[1]江西应用技术职业学院,江西赣州341000

出　　处：《电脑与电信》2020年第11期39-43,共5页Computer & Telecommunication

基　　金：基于5G的毫米波混合波束成形算法研究,项目编号:GJJ204910;赣州市科技重大专项,项目编号:[2018]50

摘　　要：针对真实场景中的噪声图像处理,根据更接近真实场景中的噪声模型,提出了一种基于卷积神经网络的去噪算法,该算法利用卷积神经网络中多个卷积层学习真实场景下噪声图像的数据特征,从而不断优化自身参数。仿真结果表明,本文所提基于卷积神经网络的去噪算法对真实场景中的噪声图像具有较好的去噪效果,去噪后的图像更加清晰,视觉效果较好,同时图像中的边缘细节得到了较好的保持。For dealing with the noise image processing in the real scene,this paper proposes a denoising algorithm based on convolutional neural network according to the noise model which closer to the real scene.The algorithm uses multiple convolution layers in the convolutional neural network to learn the data characteristics of the noise image in the real scene,so as to continuously optimize its own parameters.The simulation results show that the denoising algorithm based on convolutional neural network has a good denoising effect on the noise image in real scene,the denoised image is clearer,the visual effect is better,and the edge details in the image are well preserved.

关 键 词：图像处理 真实场景 卷积神经网络 图像去噪 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]