Color Image Restoration Exploiting Inter-Channel Correlation With a 3-Stage CNN

人工智能 脱模 计算机科学 频道(广播) 图像复原 计算机视觉 彩色图像 卷积神经网络 图像质量 特征(语言学) 像素 颜色深度 模式识别(心理学) 图像(数学) 图像处理 电信 语言学 哲学
作者
Kai Cui,Atanas Boev,Elena Alshina,Eckehard Steinbach
出处
期刊:IEEE Journal of Selected Topics in Signal Processing [Institute of Electrical and Electronics Engineers]
卷期号:15 (2): 174-189 被引量:16
标识
DOI:10.1109/jstsp.2020.3043148
摘要

Image restoration is a critical component of image processing pipelines and for low-level computer vision tasks. Conventional image restoration approaches are mostly based on hand-crafted image priors. The inter-channel correlation of color images is not fully exploited. Motivated by the special characteristics of the inter-channel correlation (higher correlation for red/green and green/blue channels than for red/blue) in color images and general characteristics (green channel always shows the best image quality among the three color components) of distorted color images, in this paper, a three-stage convolutional neural network (CNN) structure is proposed for color image restoration tasks. Since the green channel is found to have the best quality among all three channels, in the first stage, the network is designed to reconstruct the green component. Then, with the guidance of the reconstructed green channel from the first stage, the red and blue channels are reconstructed in the second stage with two parallel networks. Finally, the intermediate reconstructions from the previous stages are concatenated and further refined jointly. We demonstrate the capabilities of the proposed three-stage structure with three typical color image restoration tasks: color image demosaicking, color compression artifacts reduction, and real-world color image denoising. In addition, we integrate pixel-shuffle convolution into our scheme to improve the efficiency, and also introduce a quality-blind training strategy to simplify the training process for the compression artifacts reduction task. Extensive experimental results and analyses show that the proposed structure successfully exploits the spatial and inter-channel correlation of color images and outperforms the state-of-the-art image reconstruction approaches.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
刚刚
花成花发布了新的文献求助10
1秒前
2秒前
2秒前
自信乐安完成签到,获得积分10
2秒前
3秒前
EarholeDoctor完成签到,获得积分10
3秒前
莫誓发布了新的文献求助10
3秒前
yz发布了新的文献求助10
3秒前
4秒前
ODD完成签到,获得积分10
5秒前
5秒前
Rencal发布了新的文献求助10
6秒前
Hello应助Noel采纳,获得20
6秒前
7秒前
hanhan299发布了新的文献求助10
7秒前
灵巧的翠风完成签到 ,获得积分10
7秒前
苏漓洺发布了新的文献求助10
7秒前
汤博森发布了新的文献求助10
7秒前
8秒前
852应助清秀蝴蝶采纳,获得10
8秒前
大模型应助AoAoo采纳,获得10
8秒前
田様应助羽言采纳,获得10
8秒前
xiong完成签到 ,获得积分10
9秒前
安仔发布了新的文献求助10
9秒前
我的白起是国服完成签到 ,获得积分10
10秒前
坦率完成签到,获得积分10
10秒前
11秒前
酷酷发布了新的文献求助10
11秒前
免疫方舟完成签到,获得积分10
12秒前
12秒前
12秒前
自由的蛋挞完成签到,获得积分20
12秒前
13秒前
13秒前
14秒前
univ完成签到,获得积分10
14秒前
左囧完成签到,获得积分10
14秒前
15秒前
高分求助中
The ACS Guide to Scholarly Communication 2500
Microlepidoptera Palaearctica, Volumes 1 and 3 - 13 (12-Volume Set) [German] 1122
The Data Economy: Tools and Applications 1000
PraxisRatgeber Mantiden., faszinierende Lauerjäger. – Buch gebraucht kaufen 700
Mantiden - Faszinierende Lauerjäger – Buch gebraucht kaufen 700
Ethnicities: Media, Health, and Coping 700
书名《抗体药物研发》 560
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3092869
求助须知:如何正确求助?哪些是违规求助? 2744826
关于积分的说明 7582457
捐赠科研通 2396101
什么是DOI,文献DOI怎么找? 1270838
科研通“疑难数据库(出版商)”最低求助积分说明 614964
版权声明 598844