摘要
针对运动模糊图像的恢复,提出了基于生成式对抗神经网络(GAN)网络与FSRCNN网络的方案。采用GoPro模糊图像数据集与DIV2K图像恢复数据集进行网络训练;通过图像处理对模糊图像进行规格化、归一化、色彩空间转换等预处理操作;利用GAN对模糊图像进行图像恢复,并结合FSRCNN针对模糊图像进行图像增强;针对GAN与FSRCNN的处理结果进行分析与对比,经由FSRCNN网络图像增强的恢复图像的峰值信噪比虽然有一定的下降,但结构相似度则得到了提升。实验结果表明本文的算法方案具有较好的可行性。
An image restoration scheme based on generative adversarial networks(GAN)network and FSRCNN network for motion blurred images is proposed.GoPro fuzzy image dataset and the DIV2K image restoration dataset are used for network training.Image processing operation is used to normalize the fuzzy image,convert color space.GAN is used to restore the image of the blurred image,and the FSRCNN is combined to enhance the image for the blurred image.Processing results of GAN and FSRCNN are analyzed and compared.The peak signal-noise ratio of the reconstructed image enhanced by the FSRCNN network image has a certain decline.However,the structural similarity is improved.Experimental results show that the proposed algorithm has good feasibility.
作者
陈英
洪晨丰
CHEN Ying;HONG Chenfeng(School of Software,Nanchang Hangkong University,Nanchang 330063,China)
出处
《传感器与微系统》
CSCD
北大核心
2021年第4期63-65,73,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(61762067)。
关键词
生成式对抗神经网络
卷积神经网络
运动模糊
图像处理
generative adversarial networks(GAN)
convolutional neural network(CNN)
motion blur
image processing
