林珍玉,
崔燕,
刘蕊蕊,
杨晓辉
南昌大学信息工程学院 南昌 330031
基金项目:国家自然科学基金(61662044, 61163023)，江西省自然科学基金(20171BAB202017)
详细信息
作者简介:徐少平：男，1976年生，博士，教授，博士生导师，主要研究方向为图形图像处理技术、机器视觉、虚拟手术仿真
林珍玉：女，1996年生，硕士生，研究方向为图形图像处理技术、机器视觉
崔燕：女，1996年生，硕士生，研究方向为图形图像处理技术、机器视觉
刘蕊蕊：女，1995年生，硕士生，研究方向为图形图像处理技术、机器视觉
杨晓辉：男，1978年生，博士，副教授，主要研究方向为故障诊断及图像处理
通讯作者:徐少平 xushaoping@ncu.edu.cn
中图分类号:TN911.73; TP391计量
文章访问数:941
HTML全文浏览量:226
PDF下载量:41
被引次数:0
出版历程
收稿日期:2019-10-16
修回日期:2020-07-20
网络出版日期:2020-07-30
刊出日期:2020-10-13
A Dual-Channel Deep Convolutional Neural Network Model for Random-Valued Impulse Noise Removal
Shaoping XU,,Zhenyu LIN,
Yan CUI,
Ruirui LIU,
Xiaohui YANG
School of Information Engineering, Nanchang University, Nanchang 330031, China
Funds:The National Natural Science Foundation of China (61662044, 61163023), The Natural Science Foundation of Jiangxi Province (20171BAB202017)
摘要
摘要:为提高对随机脉冲噪声(RVIN)图像的降噪效果，该文提出一种被称为双通道降噪卷积神经网络(D-DnCNN)的RVIN深度降噪模型。首先，提取多个不同阶对数差值排序(ROLD)统计值及1个边缘特征统计值构成描述图块中心像素点是否为RVIN噪声的噪声感知特征矢量。其次，利用预先训练好的深度置信网络(DBN)预测模型实现特征矢量到噪声标签的映射，完成对噪声图像中噪声点的检测。再次，在噪声检测标签的指示下采用Delaunay三角剖分插值算法快速修复噪声像素点从而获得初步复原图像。最后，将初步复原图像作为参考图像与噪声图像联接(concatenate)后输入D-DnCNN模型后获得残差图像，将参考图像减去残差图像即可获得降噪后图像。实验数据表明：D-DnCNN模型在各个噪声比例下的降噪效果均显著超过了现有的经典开关型RVIN降噪算法，与普通的单通道RVIN深度降噪模型相比也有较大幅度提升。
关键词:图像处理/
随机脉冲噪声/
双通道降噪卷积神经网络/
参考图像/
噪声感知特征/
噪声检测/
插值
Abstract:A Dual-channel Denoising Convolutional Neural Network (D-DnCNN) model for the removal of Random-Valued Impulse Noise (RVIN) is proposed. To obtain the reference image quickly, several Rank-Ordered Logarithmic absolute Difference (ROLD) statistics and one edge feature statistic are first extracted from a local window to construct a RVIN-aware feature vector which can describe the central pixel of the patch is RVIN or not. Next, a noise detector based on Deep Belief Network (DBN) is trained to map the extracted feature vectors to their corresponding noise labels to detect all noise-like pixels in the observed image. Then, under the guidance of noise labels, the Delaunay triangulation-based interpolation algorithm is exploited to restore all detected noise-like pixels quickly and generate a preliminary restored image used as reference image. Finally, the reference image and the noisy image are simultaneously fed into the D-DnCNN model to output its corresponding residual image, and the final restored image can be obtained by subtracting the residual image from the noisy image. Extensive experimental results show that, the denoising effect of the proposed D-DnCNN denoising model outperforms the existing state-of-art switching ones across a range of noise ratios, and it also works better than the ordinary single-channel DnCNN model.
Key words:Image processing/
Random-Valued Impulse Noise(RVIN)/
Dual-channel Denoising Convolutional Neural Network(D-DnCNN)/
Reference image/
Noise-aware feature/
Noise detection/
Interpolation
PDF全文下载地址:
https://jeit.ac.cn/article/exportPdf?id=a537abe3-01d5-46ea-a831-b2edf2964232
