Multi‐scale feature fusion pyramid attention network for single image dehazing

Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide mor...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IET image processing Ročník 17; číslo 9; s. 2726 - 2735
Hlavní autoři: Liu, Jianlei, Liu, Peng, Zhang, Yuanke
Médium: Journal Article
Jazyk:angličtina
Vydáno: Wiley 01.07.2023
Témata:
ISSN:1751-9659, 1751-9667
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing is proposed. In PAN, combined with the attention mechanism, a shallow and deep feature fusion (SDF) strategy is designed. SDF considers multi‐scale as well as channel‐level fusion to provide feature information under different receptive fields while also highlighting important channels, such as texture and color information. DC is designed as a latent space mapping module to learn a mapping relationship between the latent space representation of the hazy image at low resolution and the corresponding latent space representation of the haze‐free image. Additionally, network deconvolution (ND) and deformed convolution network (DCN) are introduced into PAN. The ND module can remove pixel‐wise and channel‐wise correlation of features, reduce data redundancy to obtain sparse representation of features, and speed up network convergence. The DCN module can use its adaptive receptive field to focus on the area of interest for calculation and play a role in texture feature enhancement. Finally, the perceptual loss is chosen as the regularization item of the loss function, which makes style features of the restored image closer to the real fog‐free image. Extensive experiments reveal that the proposed PAN outperforms other existing dehazing methods on real‐world and synthetic datasets. Texture and color distortion are common in existing learning‐based dehazing algorithms, and we argue that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, we propose a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing.
AbstractList Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing is proposed. In PAN, combined with the attention mechanism, a shallow and deep feature fusion (SDF) strategy is designed. SDF considers multi‐scale as well as channel‐level fusion to provide feature information under different receptive fields while also highlighting important channels, such as texture and color information. DC is designed as a latent space mapping module to learn a mapping relationship between the latent space representation of the hazy image at low resolution and the corresponding latent space representation of the haze‐free image. Additionally, network deconvolution (ND) and deformed convolution network (DCN) are introduced into PAN. The ND module can remove pixel‐wise and channel‐wise correlation of features, reduce data redundancy to obtain sparse representation of features, and speed up network convergence. The DCN module can use its adaptive receptive field to focus on the area of interest for calculation and play a role in texture feature enhancement. Finally, the perceptual loss is chosen as the regularization item of the loss function, which makes style features of the restored image closer to the real fog‐free image. Extensive experiments reveal that the proposed PAN outperforms other existing dehazing methods on real‐world and synthetic datasets.
Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing is proposed. In PAN, combined with the attention mechanism, a shallow and deep feature fusion (SDF) strategy is designed. SDF considers multi‐scale as well as channel‐level fusion to provide feature information under different receptive fields while also highlighting important channels, such as texture and color information. DC is designed as a latent space mapping module to learn a mapping relationship between the latent space representation of the hazy image at low resolution and the corresponding latent space representation of the haze‐free image. Additionally, network deconvolution (ND) and deformed convolution network (DCN) are introduced into PAN. The ND module can remove pixel‐wise and channel‐wise correlation of features, reduce data redundancy to obtain sparse representation of features, and speed up network convergence. The DCN module can use its adaptive receptive field to focus on the area of interest for calculation and play a role in texture feature enhancement. Finally, the perceptual loss is chosen as the regularization item of the loss function, which makes style features of the restored image closer to the real fog‐free image. Extensive experiments reveal that the proposed PAN outperforms other existing dehazing methods on real‐world and synthetic datasets. Texture and color distortion are common in existing learning‐based dehazing algorithms, and we argue that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, we propose a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing.
Abstract Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow features of fog images are underutilized, and the deep features of fog images are insufficient for single image dehazing. In order to provide more texture and color information for image restoration, more shallow features need to be added in the process of image decoding. Therefore, a multi‐scale feature fusion pyramid attention network (PAN) for single image dehazing is proposed. In PAN, combined with the attention mechanism, a shallow and deep feature fusion (SDF) strategy is designed. SDF considers multi‐scale as well as channel‐level fusion to provide feature information under different receptive fields while also highlighting important channels, such as texture and color information. DC is designed as a latent space mapping module to learn a mapping relationship between the latent space representation of the hazy image at low resolution and the corresponding latent space representation of the haze‐free image. Additionally, network deconvolution (ND) and deformed convolution network (DCN) are introduced into PAN. The ND module can remove pixel‐wise and channel‐wise correlation of features, reduce data redundancy to obtain sparse representation of features, and speed up network convergence. The DCN module can use its adaptive receptive field to focus on the area of interest for calculation and play a role in texture feature enhancement. Finally, the perceptual loss is chosen as the regularization item of the loss function, which makes style features of the restored image closer to the real fog‐free image. Extensive experiments reveal that the proposed PAN outperforms other existing dehazing methods on real‐world and synthetic datasets.
Author Liu, Jianlei
Zhang, Yuanke
Liu, Peng
Author_xml – sequence: 1
  givenname: Jianlei
  surname: Liu
  fullname: Liu, Jianlei
  organization: Qufu Normal University
– sequence: 2
  givenname: Peng
  orcidid: 0000-0002-7906-3009
  surname: Liu
  fullname: Liu, Peng
  organization: Qufu Normal University
– sequence: 3
  givenname: Yuanke
  surname: Zhang
  fullname: Zhang, Yuanke
  email: yuankezhang@163.com
  organization: Qufu Normal University
BookMark eNp9kM1KA0EQhAdRMFEvPsGehej8Z_cowZ9AxCB6HnpmeuPEzW6YnSDx5CP4jD6JGyM5iHiqpqj6aKpP9uumRkJOGT1nVBYXYRn5OeM5F3ukx4aKDQqth_u7WxWHpN-2c0pVQXPVI9O7VZXC5_tH66DCrERIq9jpqg1NnS3XERbBZ5AS1mnj1Jhem_iSlU3M2lDPuk5YwAwzj8_w1hnH5KCEqsWTHz0iT9dXj6PbweT-Zjy6nAycZFIMtOMehULmKUWlvWKlkh55qRgf2qLEXFmkOh9Kq63XnnF0hfLWAqOMUiuOyHjL9Q3MzTJ2X8S1aSCYb6OJMwMxBVehARDaCpHnwmsJmkKOjAsnuXRUSYsd62zLcrFp24jljseo2exqNrua7127MP0VdiHBZpwUIVR_V9i28hoqXP8DN-PpA992vgCLiY57
CitedBy_id crossref_primary_10_1007_s00371_023_03177_2
crossref_primary_10_3390_s25123750
crossref_primary_10_1049_ipr2_70043
Cites_doi 10.1007/978-3-642-33715-4_54
10.1109/CVPRW50498.2020.00230
10.1109/CVPR.2016.90
10.1109/ICCV.2019.00254
10.1109/CVPR.2017.106
10.1016/j.patcog.2020.107255
10.1109/ICCV.2017.511
10.1023/A:1016328200723
10.1007/s11263-015-0816-y
10.1145/3478457
10.1109/ICICSE55337.2022.9828891
10.1109/TMM.2022.3155937
10.1109/TPAMI.2018.2882478
10.1109/TIP.2018.2867951
10.1063/1.3037551
10.1109/ICCV.2019.00337
10.1109/CVPR46437.2021.01041
10.1109/TGRS.2022.3204890
10.1007/978-3-319-46475-6_10
10.1049/ipr2.12455
10.1109/WACV.2019.00151
10.1016/j.sigpro.2018.03.008
10.1109/TIP.2022.3214093
10.1049/ipr2.12467
10.1109/TIP.2021.3060873
10.1109/TIP.2015.2446191
10.1007/978-3-319-46475-6_43
10.1109/CVPR.2014.383
10.1109/TIP.2016.2598681
10.1109/CVPR.2018.00745
10.1007/s00521‐021‐06296‐w
10.1109/CVPR52688.2022.01167
10.1109/ICCV.2017.89
ContentType Journal Article
Copyright 2023 The Authors. published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology.
Copyright_xml – notice: 2023 The Authors. published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology.
DBID 24P
AAYXX
CITATION
DOA
DOI 10.1049/ipr2.12823
DatabaseName Wiley Online Library Open Access
CrossRef
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
DatabaseTitleList CrossRef


Database_xml – sequence: 1
  dbid: 24P
  name: Wiley Online Library Open Access
  url: https://authorservices.wiley.com/open-science/open-access/browse-journals.html
  sourceTypes: Publisher
– sequence: 2
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISSN 1751-9667
EndPage 2735
ExternalDocumentID oai_doaj_org_article_aa36b33883d64a60a8e123c424c054be
10_1049_ipr2_12823
IPR212823
Genre article
GrantInformation_xml – fundername: National Natural Science Foundation of China
  funderid: 61773243
GroupedDBID .DC
0R~
1OC
24P
29I
5GY
6IK
8VB
AAHHS
AAHJG
AAJGR
ABQXS
ACCFJ
ACCMX
ACESK
ACGFS
ACIWK
ACXQS
ADZOD
AEEZP
AENEX
AEQDE
AIWBW
AJBDE
ALMA_UNASSIGNED_HOLDINGS
ALUQN
AVUZU
CS3
DU5
EBS
ESX
GROUPED_DOAJ
HZ~
IAO
IFIPE
IPLJI
ITC
JAVBF
K1G
LAI
MCNEO
MS~
O9-
OCL
OK1
P2P
QWB
RIE
RNS
ROL
RUI
ZL0
4.4
8FE
8FG
AAMMB
AAYXX
ABJCF
AEFGJ
AFFHD
AFKRA
AGXDD
AIDQK
AIDYY
ARAPS
BENPR
BGLVJ
CCPQU
CITATION
EJD
HCIFZ
IDLOA
L6V
M43
M7S
P62
PHGZM
PHGZT
PQGLB
PTHSS
S0W
WIN
ID FETCH-LOGICAL-c4143-6c2de35e1d00e56d51f54de2f5127b9fe85be06874b6bd6d12ec95dbba10100b3
IEDL.DBID DOA
ISICitedReferencesCount 3
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000985022400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1751-9659
IngestDate Fri Oct 03 12:37:02 EDT 2025
Wed Oct 29 21:25:23 EDT 2025
Tue Nov 18 22:33:50 EST 2025
Wed Jan 22 16:19:24 EST 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 9
Language English
License Attribution-NonCommercial-NoDerivs
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c4143-6c2de35e1d00e56d51f54de2f5127b9fe85be06874b6bd6d12ec95dbba10100b3
ORCID 0000-0002-7906-3009
OpenAccessLink https://doaj.org/article/aa36b33883d64a60a8e123c424c054be
PageCount 10
ParticipantIDs doaj_primary_oai_doaj_org_article_aa36b33883d64a60a8e123c424c054be
crossref_primary_10_1049_ipr2_12823
crossref_citationtrail_10_1049_ipr2_12823
wiley_primary_10_1049_ipr2_12823_IPR212823
PublicationCentury 2000
PublicationDate 2023-07-01
PublicationDateYYYYMMDD 2023-07-01
PublicationDate_xml – month: 07
  year: 2023
  text: 2023-07-01
  day: 01
PublicationDecade 2020
PublicationTitle IET image processing
PublicationYear 2023
Publisher Wiley
Publisher_xml – name: Wiley
References 2010; 33
2018; 28
2023; 35
2012
2018; 149
2020; 34
2020; 102
2021; 30
2018; 42
2015; 24
2002; 48
2015; 115
2022
2022; 60
2021
2020
2019
2018
1977; 30
2017
2016
2022; 31
2014
2016; 25
2022; 16
2022; 18
e_1_2_9_30_1
e_1_2_9_31_1
e_1_2_9_11_1
Qin X. (e_1_2_9_19_1) 2020; 34
e_1_2_9_34_1
e_1_2_9_10_1
e_1_2_9_35_1
e_1_2_9_13_1
e_1_2_9_32_1
e_1_2_9_12_1
e_1_2_9_33_1
Xu X. (e_1_2_9_27_1) 2019
e_1_2_9_15_1
e_1_2_9_38_1
e_1_2_9_14_1
e_1_2_9_39_1
e_1_2_9_17_1
e_1_2_9_36_1
e_1_2_9_16_1
e_1_2_9_37_1
e_1_2_9_18_1
e_1_2_9_20_1
e_1_2_9_40_1
e_1_2_9_22_1
e_1_2_9_21_1
He K. (e_1_2_9_6_1) 2010; 33
e_1_2_9_24_1
e_1_2_9_23_1
e_1_2_9_8_1
e_1_2_9_7_1
e_1_2_9_5_1
e_1_2_9_4_1
e_1_2_9_3_1
e_1_2_9_2_1
e_1_2_9_9_1
e_1_2_9_26_1
e_1_2_9_25_1
e_1_2_9_28_1
e_1_2_9_29_1
References_xml – start-page: 3276
  year: 2019
  end-page: 3285
  article-title: LAP‐Net: Level‐aware progressive network for image dehazing
– start-page: 105
  year: 2022
  end-page: 109
  article-title: MARG‐UNet: A Single Image Dehazing Network Based on Multimodal Attention Residual Group
– volume: 24
  start-page: 3522
  issue: 11
  year: 2015
  end-page: 3533
  article-title: A fast single image haze removal algorithm using color attenuation prior
  publication-title: IEEE Trans. Image Process.
– start-page: 11976
  year: 2022
  end-page: 11986
  article-title: A convnet for the 2020s
– volume: 149
  start-page: 135
  year: 2018
  end-page: 147
  article-title: Image dehazing by artificial multiple‐exposure image fusion
  publication-title: Signal Process
– volume: 60
  start-page: 1
  year: 2022
  end-page: 13
  article-title: Dehaze‐AGGAN: Unpaired Remote Sensing Image Dehazing Using Enhanced Attention‐Guide Generative Adversarial Networks
  publication-title: IEEE Trans. Geosci. Remote Sensing
– start-page: 154
  year: 2016
  end-page: 169
  article-title: Single image dehazing via multi‐scale convolutional neural networks
– start-page: 490
  year: 2020
  end-page: 491
  article-title: Ntire 2020 challenge on nonhomogeneous dehazing
– volume: 34
  start-page: 11908
  issue: 07
  year: 2020
  end-page: 11915
  article-title: FFA‐Net: Feature fusion attention network for single image dehazing
  publication-title: Proc. AAAI Conf. Artif. Intell.
– start-page: 2117
  year: 2017
  end-page: 2125
  article-title: Feature pyramid networks for object detection
– volume: 42
  start-page: 720
  issue: 3
  year: 2018
  end-page: 734
  article-title: Single image dehazing using haze‐lines
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 31
  start-page: 6635
  year: 2022
  end-page: 6648
  article-title: BiN‐Flow: Bidirectional Normalizing Flow for Robust Image Dehazing
  publication-title: IEEE Trans. Image Process.
– year: 2022
  article-title: MSAFF‐Net: Multiscale Attention Feature Fusion Networks for Single Image Dehazing and Beyond
  publication-title: IEEE Trans. Multim.
– volume: 25
  start-page: 5187
  issue: 11
  year: 2016
  end-page: 5198
  article-title: Dehazenet: An end‐to‐end system for single image haze removal
  publication-title: IEEE Trans. Image Process.
– start-page: 10551
  year: 2021
  end-page: 10560
  article-title: Contrastive learning for compact single image dehazing
– start-page: 7132
  year: 2018
  end-page: 7141
  article-title: Squeeze‐and‐excitation networks
– start-page: 770
  year: 2016
  end-page: 778
  article-title: Deep residual learning for image recognition
– start-page: 2995
  year: 2014
  end-page: 3000
  article-title: Investigating haze‐relevant features in a learning framework for image dehazing
– year: 2014
– start-page: 1375
  year: 2019
  end-page: 1383
  article-title: Gated context aggregation network for image dehazing and deraining
– volume: 30
  start-page: 3391
  year: 2021
  end-page: 3404
  article-title: RefineDNet: A weakly supervised refinement framework for single image dehazing
  publication-title: IEEE Trans. Image Process.
– start-page: 444
  year: 2020
  end-page: 445
  article-title: NH‐HAZE: An image dehazing benchmark with non‐homogeneous hazy and haze‐free images
– volume: 16
  start-page: 2049
  issue: 8
  year: 2022
  end-page: 2062
  article-title: Multi‐scale single image dehazing based on the fusion of global and local features
  publication-title: IET Image Process.
– start-page: 746
  year: 2012
  end-page: 760
  article-title: Indoor segmentation and support inference from rgbd images
– volume: 18
  issue: 2
  year: 2022
  article-title: SADnet: Semi‐supervised Single Image Dehazing Method Based on an Attention Mechanism
  publication-title: ACM Trans. Multimedia Comput. Commun. Appl.
– volume: 115
  start-page: 211
  issue: 3
  year: 2015
  end-page: 252
  article-title: Imagenet large scale visual recognition challenge
  publication-title: Int. J. Comput. Vis.
– start-page: 764
  year: 2017
  end-page: 773
  article-title: Deformable convolutional networks
– start-page: 4770
  year: 2017
  end-page: 4778
  article-title: Aod‐net: All‐in‐one dehazing network
– volume: 33
  start-page: 2341
  issue: 12
  year: 2010
  end-page: 2353
  article-title: Single image haze removal using dark channel prior
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– year: 2019
  article-title: Learning deformable kernels for image and video denoising
  publication-title: ArXiv Prepr
– volume: 35
  start-page: 3737
  issue: 5
  year: 2023
  end-page: 3751
  article-title: Scale‐free heterogeneous cycleGAN for defogging from a single image for autonomous driving in fog
  publication-title: Neural Comp. Appl.
– start-page: 2453
  year: 2019
  end-page: 2462
  article-title: Deep multi‐model fusion for single‐image dehazing
– volume: 48
  start-page: 233
  issue: 3
  year: 2002
  end-page: 254
  article-title: Vision and the atmosphere
  publication-title: Int. J. Comput. Vis.
– volume: 16
  start-page: 1897
  issue: 7
  year: 2022
  end-page: 1907
  article-title: Single image dehazing using generative adversarial networks based on an attention mechanism
  publication-title: IET Image Process.
– volume: 102
  year: 2020
  article-title: A novel image‐dehazing network with a parallel attention block
  publication-title: Pattern Recognit
– volume: 28
  start-page: 492
  issue: 1
  year: 2018
  end-page: 505
  article-title: Benchmarking single‐image dehazing and beyond
  publication-title: IEEE Trans. Image Process.
– year: 2019
– start-page: 694
  year: 2016
  end-page: 711
  article-title: Perceptual losses for real‐time style transfer and super‐resolution
– volume: 30
  start-page: 76
  issue: 5
  year: 1977
  end-page: 77
  article-title: Optics of the atmosphere: Scattering by molecules and particles
  publication-title: Phys. Today
– ident: e_1_2_9_29_1
  doi: 10.1007/978-3-642-33715-4_54
– ident: e_1_2_9_30_1
  doi: 10.1109/CVPRW50498.2020.00230
– ident: e_1_2_9_21_1
  doi: 10.1109/CVPR.2016.90
– ident: e_1_2_9_9_1
  doi: 10.1109/ICCV.2019.00254
– ident: e_1_2_9_31_1
– ident: e_1_2_9_25_1
– ident: e_1_2_9_13_1
  doi: 10.1109/CVPR.2017.106
– ident: e_1_2_9_15_1
  doi: 10.1016/j.patcog.2020.107255
– ident: e_1_2_9_32_1
  doi: 10.1109/ICCV.2017.511
– ident: e_1_2_9_4_1
  doi: 10.1023/A:1016328200723
– ident: e_1_2_9_26_1
  doi: 10.1007/s11263-015-0816-y
– ident: e_1_2_9_36_1
  doi: 10.1145/3478457
– ident: e_1_2_9_38_1
  doi: 10.1109/ICICSE55337.2022.9828891
– ident: e_1_2_9_37_1
  doi: 10.1109/TMM.2022.3155937
– ident: e_1_2_9_5_1
  doi: 10.1109/TPAMI.2018.2882478
– ident: e_1_2_9_28_1
  doi: 10.1109/TIP.2018.2867951
– ident: e_1_2_9_3_1
  doi: 10.1063/1.3037551
– ident: e_1_2_9_8_1
  doi: 10.1109/ICCV.2019.00337
– ident: e_1_2_9_20_1
  doi: 10.1109/CVPR46437.2021.01041
– ident: e_1_2_9_39_1
  doi: 10.1109/TGRS.2022.3204890
– ident: e_1_2_9_12_1
  doi: 10.1007/978-3-319-46475-6_10
– volume: 34
  start-page: 11908
  issue: 07
  year: 2020
  ident: e_1_2_9_19_1
  article-title: FFA‐Net: Feature fusion attention network for single image dehazing
  publication-title: Proc. AAAI Conf. Artif. Intell.
– ident: e_1_2_9_16_1
  doi: 10.1049/ipr2.12455
– ident: e_1_2_9_18_1
  doi: 10.1109/WACV.2019.00151
– ident: e_1_2_9_33_1
  doi: 10.1016/j.sigpro.2018.03.008
– ident: e_1_2_9_35_1
  doi: 10.1109/TIP.2022.3214093
– ident: e_1_2_9_17_1
  doi: 10.1049/ipr2.12467
– ident: e_1_2_9_34_1
  doi: 10.1109/TIP.2021.3060873
– year: 2019
  ident: e_1_2_9_27_1
  article-title: Learning deformable kernels for image and video denoising
  publication-title: ArXiv Prepr
– ident: e_1_2_9_7_1
  doi: 10.1109/TIP.2015.2446191
– volume: 33
  start-page: 2341
  issue: 12
  year: 2010
  ident: e_1_2_9_6_1
  article-title: Single image haze removal using dark channel prior
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– ident: e_1_2_9_24_1
  doi: 10.1007/978-3-319-46475-6_43
– ident: e_1_2_9_10_1
  doi: 10.1109/CVPR.2014.383
– ident: e_1_2_9_11_1
  doi: 10.1109/TIP.2016.2598681
– ident: e_1_2_9_23_1
– ident: e_1_2_9_2_1
  doi: 10.1109/CVPR.2018.00745
– ident: e_1_2_9_40_1
  doi: 10.1007/s00521‐021‐06296‐w
– ident: e_1_2_9_14_1
  doi: 10.1109/CVPR52688.2022.01167
– ident: e_1_2_9_22_1
  doi: 10.1109/ICCV.2017.89
SSID ssj0059085
Score 2.3174055
Snippet Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the shallow...
Abstract Texture and color distortion are common in existing learning‐based dehazing algorithms, and it is argued that one of the major reasons is that the...
SourceID doaj
crossref
wiley
SourceType Open Website
Enrichment Source
Index Database
Publisher
StartPage 2726
SubjectTerms attention mechanism
feature fusion
image dehazing
pyramid autoencoder
SummonAdditionalLinks – databaseName: Wiley Online Library Open Access
  dbid: 24P
  link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1dS91AEB2s9aEvta1Kr_1gob4oRJPdzSaBvrSl0r7IRSr4FvZjVi_U6yVXC33rT_A39pd0ZpN7RRBB-hbChISdnZmzm51zAHZKHbju0rIkxirjVs2sUS7PClTeedUUWMckNlEdHdWnp814BT4uemF6fojlhhtHRsrXHODW9SokBGrJiZNZJ_cpu0r1BJ4Whap4Tks9XuRhFvMuUzskC8mbslmQk-rm4PbZO-UosfbfRampzByu_98HvoDnA7wUn_r58BJWcPoK1geoKYZAnm_AODXe_v1zMycfoYiY-D1FvObNMzH73dmLSRDMvZlOQ4ppf1pcEMQVvLtAz0wuKBWJgOdMUH22CSeHX398-ZYN4gqZ14SRMuNlQFViEfIcSxPKIpLfUEZCAJVrItalw9zUlXbGBRMKib4pg3OWgjjPndqC1enlFF-DiJZAm7aWmWY0ARZyMC1iUCmb6yB1GMHuYoxbPzCPswDGzzb9AddNy0PVpqEawYel7azn27jX6jO7amnBHNnpxmV31g4h11qrjKMVeK2C0dbktkYq015L7QmnOhzBXnLfA-9pv4-PZbrafozxG3jGqvT9qd63sHrVXeM7WPO_ribz7n2apf8AqX3p2Q
  priority: 102
  providerName: Wiley-Blackwell
Title Multi‐scale feature fusion pyramid attention network for single image dehazing
URI https://onlinelibrary.wiley.com/doi/abs/10.1049%2Fipr2.12823
https://doaj.org/article/aa36b33883d64a60a8e123c424c054be
Volume 17
WOSCitedRecordID wos000985022400001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 1751-9667
  dateEnd: 20241231
  omitProxy: false
  ssIdentifier: ssj0059085
  issn: 1751-9659
  databaseCode: DOA
  dateStart: 20210101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVWIB
  databaseName: Wiley Online Library Free Content
  customDbUrl:
  eissn: 1751-9667
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0059085
  issn: 1751-9659
  databaseCode: WIN
  dateStart: 20130101
  isFulltext: true
  titleUrlDefault: https://onlinelibrary.wiley.com
  providerName: Wiley-Blackwell
– providerCode: PRVWIB
  databaseName: Wiley Online Library Open Access
  customDbUrl:
  eissn: 1751-9667
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0059085
  issn: 1751-9659
  databaseCode: 24P
  dateStart: 20130101
  isFulltext: true
  titleUrlDefault: https://authorservices.wiley.com/open-science/open-access/browse-journals.html
  providerName: Wiley-Blackwell
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV1NT9wwELUAcegFWqBioSBL5QJSuo7tOPERqq7KZbVCILhF_hjDSuyy2g-k3vgJ_EZ-ScdOFhWpKpdeoihylGgmnnnjjN8j5KiQPuZdLEtCKLO4VTPTwrIsB-GsEzqHKiSxibLfr25u9OAPqa_YE9bQAzeG6xojlMU6qhJeSaOYqQCDrZNcOkQbFmL0ZaVeFlNNDI5C3kXaChlF5FWhl8SkUneHkyn_hlGZizepKDH2v0WoKcX0PpKNFhvS0-adPpEVGG-RzRYn0nYWzrbJIO2afXl6nqGBgQZI5Jw0LOLKF538mprR0NNInJlaGem4afWmiE9pXBrAe4YjjCPUw11kl77dIVe9H5fff2atMkLmJAKcTDnuQRSQe8agUL7IAxodeMD0XVodoCosMFWV0irrlc85OF14aw3OQMas-EzWxg9j2CU0GERc0phIEyMRbaB3sAIBIQyTnkvfIcdLI9WupQ2P6hX3dfp9LXUdDVong3bI19exk4Ys46-jzqKtX0dEgut0Ad1et26v33N7h5wkT_3jOfX54IKns73_8cR98iFKzTetul_I2ny6gAOy7h7nw9n0kKxyOThMHyEer8_7vwGTbN8B
linkProvider Directory of Open Access Journals
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1fa9RAEB9qK-iLrf_wbK0L-qIQm-xu9pLHViwt1uOQCn0L-2e2HtjrkWsF3_wI_Yx-ks5scicFEUrfQpglYScz85vNzG8A3pY6cNyltCTGYcatmlmtXJ4VqLzzqi6wimnYxHA0qk5O6nFfm8O9MB0_xPLAjS0j-Ws2cD6Q7hJOzSSZk1krP5B7leoerGkKMzzAQOrxwhHzNO8y9UPyJHlT1gt2Ul3v_F17Ix4l2v6bMDXFmf31O77hBjzqAabY7b6Ix7CC0yew3oNN0Zvy_CmMU-vtn99Xc9ISioiJ4VPESz4-E7NfrT2bBMHsm6keUky7enFBIFfw-QKtmZyRMxIBvzNF9ekz-Lb_6fjjQdaPV8i8JpSUGS8DqhKLkOdYmlAWkTSHMhIGGLo6YlU6zE011M64YEIh0ddlcM6SGee5U89hdXo-xRcgoiXYpq1lrhlNkIVUTGkMKmVzHaQOA3i32OTG99zjPALjR5P-geu64a1q0lYN4M1SdtYxbvxTao91tZRglux047w9bXqja6xVxlEOXqlgtDW5rZACtddSe0KqDgfwPunvP89pDsdfZbp6eRvh1_Dg4PjLUXN0OPq8CQ95Rn1X47sFqxftJb6C-_7nxWTebqdP9hqLSe27
linkToPdf http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1ba9RQEB5qK-KLrTdcL_WAvihEk5zLJo9eurQoSxCFvoVzmVMX7HbJtoW--RP8jf0lzpxkVwoiSN9CmEPCmczMNycz3wC81Cpw3KW0JMZxxq2aWS1dnhUovfOyLrCKadjEeDqtDg_rZqjN4V6Ynh9ifeDGlpH8NRs4LkLsE07FJJmzRVe-IfdayhuwpTQ5WSZ2Vs3KEfM0b536IXmSvNH1ip1U1W__rL0SjxJt_1WYmuLMZPuab7gDdwaAKd71X8Rd2MD5PdgewKYYTHl5H5rUenv589eStIQiYmL4FPGMj8_E4qKzx7MgmH0z1UOKeV8vLgjkCj5foDWzY3JGIuB3pqg-egDfJntfP-xnw3iFzCtCSZnxZUCpsQh5jtoEXUTSHJaRMMDY1REr7TA31Vg544IJRYm-1sE5S2ac504-hM35yRwfgYiWYJuylrlmFEEWUjGlMSilzVUoVRjBq9Umt37gHucRGD_a9A9c1S1vVZu2agQv1rKLnnHjr1LvWVdrCWbJTjdOuqN2MLrWWmkc5eCVDEZZk9sKKVB7VSpPSNXhCF4n_f3jOe1B86VMV4__R_g53Go-TtrPB9NPT-A2j6jvS3yfwuZpd4bP4KY_P50tu930xf4GSPjtPw
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Multi%E2%80%90scale+feature+fusion+pyramid+attention+network+for+single+image+dehazing&rft.jtitle=IET+image+processing&rft.au=Liu%2C+Jianlei&rft.au=Liu%2C+Peng&rft.au=Zhang%2C+Yuanke&rft.date=2023-07-01&rft.issn=1751-9659&rft.eissn=1751-9667&rft.volume=17&rft.issue=9&rft.spage=2726&rft.epage=2735&rft_id=info:doi/10.1049%2Fipr2.12823&rft.externalDBID=n%2Fa&rft.externalDocID=10_1049_ipr2_12823
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1751-9659&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1751-9659&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1751-9659&client=summon