A data-driven algorithm for solving image despeckling PDE model using physics-informed ConvNet

In this study, we propose a new data-driven algorithm for the Perona-Malik image despeckling problem. The advantage of the proposed algorithm over neural network-based methods is that it does not need any noisy and clean image data pair for training. The proposed algorithm is implemented using a thr...

Full description

Saved in:
Bibliographic Details
Published in:Computers & mathematics with applications (1987) Vol. 200; pp. 202 - 227
Main Authors: Kumar, Haridarshan, Kumar, Sanjeev
Format: Journal Article
Language:English
Published: Elsevier Ltd 15.12.2025
Subjects:
ISSN:0898-1221
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract In this study, we propose a new data-driven algorithm for the Perona-Malik image despeckling problem. The advantage of the proposed algorithm over neural network-based methods is that it does not need any noisy and clean image data pair for training. The proposed algorithm is implemented using a three-dimensional convolution neural network (ConvNet) architecture. We compare its output with results obtained from several existing methods, including the operator splitting RBF collocation method, the finite difference method (FDM), and physics-informed neural networks (PINNs). To evaluate the performance of the proposed algorithm, simulations are carried out using grayscale images that have been artificially corrupted with different levels of speckle noise. Using the peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) as the evaluation metric, we observed that the proposed algorithm outperforms these existing methods, demonstrating superior image quality with the same numerical scheme and the same discretization. To the best of our knowledge, this work represents the first application of physics-inspired convolutional neural network for PDE-based image despeckling model.
AbstractList In this study, we propose a new data-driven algorithm for the Perona-Malik image despeckling problem. The advantage of the proposed algorithm over neural network-based methods is that it does not need any noisy and clean image data pair for training. The proposed algorithm is implemented using a three-dimensional convolution neural network (ConvNet) architecture. We compare its output with results obtained from several existing methods, including the operator splitting RBF collocation method, the finite difference method (FDM), and physics-informed neural networks (PINNs). To evaluate the performance of the proposed algorithm, simulations are carried out using grayscale images that have been artificially corrupted with different levels of speckle noise. Using the peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) as the evaluation metric, we observed that the proposed algorithm outperforms these existing methods, demonstrating superior image quality with the same numerical scheme and the same discretization. To the best of our knowledge, this work represents the first application of physics-inspired convolutional neural network for PDE-based image despeckling model.
Author Kumar, Haridarshan
Kumar, Sanjeev
Author_xml – sequence: 1
  givenname: Haridarshan
  surname: Kumar
  fullname: Kumar, Haridarshan
  email: haridarshan_k@ma.iitr.ac.in
  organization: Department of Mathematics, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, 247667, India
– sequence: 2
  givenname: Sanjeev
  surname: Kumar
  fullname: Kumar, Sanjeev
  email: sanjeev.kumar@ma.iitr.ac.in
  organization: Department of Mathematics, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, 247667, India
BookMark eNp9kMtOwzAQRb0oEm3hC9j4BxL8IKm9YFGV8pAqYAFbLMcety5JXNkhqH9PQlmzmrkjndHVmaFJG1pA6IqSnBJaXu9zo5tvnTPCipzInDA2QVMipMgoY_QczVLaE0JuOCNT9LHEVnc6s9H30GJdb0P03a7BLkScQt37dot9o7eALaQDmM96vLzerXETLNT4K435sDsmb1Lm24FrwOJVaPtn6C7QmdN1gsu_OUfv9-u31WO2eXl4Wi03mWFF0WXaEaOronIMhJNSD0tJC14CM2CYFIIJbksu7YJVFIBwWwlXCXBDXFjH-Rzx018TQ0oRnDrEoXU8KkrUqEXt1a8WNWpRRKpBy0DdnigYqvUeokrGQ2vA-gimUzb4f_kfo-NyjA
Cites_doi 10.1109/TPAMI.1980.4766994
10.1016/j.camwa.2023.10.002
10.1016/j.isprsjprs.2024.11.003
10.1109/42.802756
10.1016/0893-6080(89)90020-8
10.1006/jvci.2001.0502
10.1016/j.jcp.2018.10.045
10.1109/5.726791
10.1109/82.365355
10.1016/j.cma.2020.113028
10.1109/34.56205
10.4208/cicp.OA-2020-0164
10.1016/S0167-8655(97)00157-8
10.1016/j.camwa.2022.03.013
10.1109/4233.897062
10.1364/JOSA.66.001145
10.1007/s10851-011-0285-z
10.1007/s10409-021-01148-1
10.1109/42.938245
10.1016/j.camwa.2017.07.036
10.1016/j.cma.2020.113547
10.1109/TIP.2002.804276
10.1016/j.csda.2012.05.009
10.1109/MGRS.2013.2277512
10.1016/j.cma.2024.116813
10.1109/TPAMI.1982.4767223
10.1007/s10569-022-10069-5
10.1109/TIP.2009.2029593
10.1137/060671814
10.1109/TPAMI.1985.4767641
10.1016/j.patrec.2025.02.021
10.1007/s40096-021-00444-y
10.1109/83.869184
10.1109/TCSVT.2006.888020
10.1007/s11042-022-12998-7
10.1038/s42254-021-00314-5
10.1007/s10851-010-0225-3
10.1016/j.cnsns.2024.107911
10.1007/s11042-018-6576-2
ContentType Journal Article
Copyright 2025 Elsevier Ltd
Copyright_xml – notice: 2025 Elsevier Ltd
DBID AAYXX
CITATION
DOI 10.1016/j.camwa.2025.09.022
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EndPage 227
ExternalDocumentID 10_1016_j_camwa_2025_09_022
S0898122125004031
GroupedDBID --K
--M
-~X
.DC
.~1
0R~
1B1
1RT
1~.
1~5
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
AAEDT
AAEDW
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AATTM
AAXKI
AAXUO
AAYFN
ABAOU
ABBOA
ABJNI
ABMAC
ACDAQ
ACGFS
ACIWK
ACLOT
ACNCT
ACRLP
ACVFH
ACZNC
ADBBV
ADCNI
ADEZE
ADTZH
AEBSH
AECPX
AEIPS
AEKER
AENEX
AEUPX
AFJKZ
AFPUW
AFTJW
AGUBO
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIGII
AIGVJ
AIIUN
AIKHN
AITUG
AKBMS
AKRWK
AKYEP
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
ANKPU
AOUOD
APXCP
ARUGR
AXJTR
BJAXD
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFKBS
EFLBG
EO8
EO9
EP2
EP3
F5P
FDB
FEDTE
FIRID
FNPLU
FYGXN
G-Q
GBLVA
GBOLZ
HVGLF
IHE
IXB
J1W
JJJVA
KOM
LG9
M26
M41
MHUIS
MO0
N9A
O-L
O9-
OAUVE
OK1
OZT
P-8
P-9
P2P
PC.
PQQKQ
Q38
RNS
ROL
RPZ
SDF
SDG
SDP
SES
SEW
SPC
SPCBC
SST
SSV
SSW
T5K
TN5
XPP
ZMT
~G-
~HD
29F
9DU
AAQXK
AAYWO
AAYXX
ABFNM
ABWVN
ABXDB
ACNNM
ACRPL
ADMUD
ADNMO
ADVLN
AEXQZ
AFFNX
AGHFR
AGQPQ
ASPBG
AVWKF
AZFZN
CITATION
EJD
FGOYB
G-2
HZ~
R2-
SSZ
TAE
WUQ
ZY4
ID FETCH-LOGICAL-c255t-af0cab5bf2e8f99abf261536e2cec2988283d639d72b1ee03db8fb8ef2b17df33
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001585693000002&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0898-1221
IngestDate Thu Nov 27 00:52:52 EST 2025
Wed Dec 10 14:37:49 EST 2025
IsPeerReviewed true
IsScholarly true
Keywords Physics-informed machine learning
Convolutional neural networks
Diffusion
Image despeckling
PDEs
Finite difference method
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c255t-af0cab5bf2e8f99abf261536e2cec2988283d639d72b1ee03db8fb8ef2b17df33
PageCount 26
ParticipantIDs crossref_primary_10_1016_j_camwa_2025_09_022
elsevier_sciencedirect_doi_10_1016_j_camwa_2025_09_022
PublicationCentury 2000
PublicationDate 2025-12-15
PublicationDateYYYYMMDD 2025-12-15
PublicationDate_xml – month: 12
  year: 2025
  text: 2025-12-15
  day: 15
PublicationDecade 2020
PublicationTitle Computers & mathematics with applications (1987)
PublicationYear 2025
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Aubert, Aujol (br0040) 2008; 68
Bhuiyan, Ahmad, Swamy (br0010) 2007; 17
Nandal, Kumar (br0330) 2019; 78
Namaki, Eslahchi, Salehi (br0490) 2023; 152
Eslahchi, Esmaili, Namaki, Salehi (br0360) 2021; 17
Cai, Mao, Wang, Yin, Karniadakis (br0470) 2021; 37
Zhang, Ye (br0100) 2017; 74
Argenti, Lapini, Bianchi, Alparone (br0170) 2013; 1
Fang, Liu, Peng, Guan, Li, Tian (br0130) 2024; 218
Jin, Wang, Hiller (br0290) 2000; 4
Rudin, Lions, Osher (br0060) 2003
Frost, Stiles, Shanmugan, Holtzman (br0250) 1982; 4
Hornik, Stinchcombe, White (br0510) 1989; 2
Zhang, Wei, Xiao (br0310) 2012; 43
Gagnon, Jouan (br0020) 1997
Alparone, Baronti, Carla (br0180) 1995; 42
Shi, Zeng, Liang (br0500) 2024; 132
Alparone, Garzelli (br0190) 1998; 19
Yuan, Guan, Feng, Wu (br0140) 2020; 19
You, Kaveh (br0090) 2000; 9
Guo, Odegard, Lang, Gopinath, Selesnick, Burrus (br0030) 1994
Smith (br0370) 1985
Weickert (br0350) 1998
Jagtap, Karniadakis (br0430) 2020; 28
Karniadakis, Kevrekidis, Lu, Perdikaris, Wang, Yang (br0530) 2021; 3
Guo, Pu, Huang (br0320) 2015
Yu, Acton (br0300) 2002; 11
Teuber, Lang (br0230) 2012; 56
Süli (br0380) 2002
Jin, Yang (br0050) 2011; 39
Deledalle, Denis, Tupin (br0220) 2009; 18
Yuan, Wu, Feng, Fu, Wu (br0120) 2023; 61
Jagtap, Kharazmi, Karniadakis (br0440) 2020; 365
Hao, Gao, Gao (br0280) 1999; 18
Martin, Schaub (br0480) 2022; 134
Lee (br0270) 1980; PAMI-2
Wang, Wu, Yuan, Fu, Zhang (br0150) 2024
Prager, Gee, Treece, Berman (br0200) 2002; 415
Witkin (br0070) 1987
Kharazmi, Zhang, Karniadakis (br0450) 2021; 374
Eymard, Gallouët, Herbin (br0390) 2000; 7
Albisani, Baracchi, Piva, Argenti (br0110) 2025; 190
Lakra, Kumar (br0340) 2022; 81
Krivá, Mikula (br0400) 2002; 13
Raissi, Perdikaris, Karniadakis (br0420) 2019; 378
Perona, Malik (br0080) 1990; 12
Achim, Bezerianos, Tsakalides (br0240) 2001; 20
Wang, Sankaran, Perdikaris (br0460) 2024; 421
Lotfi, Parand (br0410) 2022; 113
Buades, Coll, Morel (br0210) 2005
Kuan, Sawchuk, Strand, Chavel (br0260) 1985; 7
LeCun, Bottou, Bengio, Haffner (br0520) 1998; 86
Goodman (br0160) 1976; 66
Wang (10.1016/j.camwa.2025.09.022_br0460) 2024; 421
Goodman (10.1016/j.camwa.2025.09.022_br0160) 1976; 66
Alparone (10.1016/j.camwa.2025.09.022_br0180) 1995; 42
Argenti (10.1016/j.camwa.2025.09.022_br0170) 2013; 1
Zhang (10.1016/j.camwa.2025.09.022_br0310) 2012; 43
Bhuiyan (10.1016/j.camwa.2025.09.022_br0010) 2007; 17
Wang (10.1016/j.camwa.2025.09.022_br0150) 2024
LeCun (10.1016/j.camwa.2025.09.022_br0520) 1998; 86
Eymard (10.1016/j.camwa.2025.09.022_br0390) 2000; 7
Fang (10.1016/j.camwa.2025.09.022_br0130) 2024; 218
Frost (10.1016/j.camwa.2025.09.022_br0250) 1982; 4
Aubert (10.1016/j.camwa.2025.09.022_br0040) 2008; 68
Hao (10.1016/j.camwa.2025.09.022_br0280) 1999; 18
Süli (10.1016/j.camwa.2025.09.022_br0380) 2002
Krivá (10.1016/j.camwa.2025.09.022_br0400) 2002; 13
Jin (10.1016/j.camwa.2025.09.022_br0050) 2011; 39
Lotfi (10.1016/j.camwa.2025.09.022_br0410) 2022; 113
Guo (10.1016/j.camwa.2025.09.022_br0030) 1994
You (10.1016/j.camwa.2025.09.022_br0090) 2000; 9
Teuber (10.1016/j.camwa.2025.09.022_br0230) 2012; 56
Smith (10.1016/j.camwa.2025.09.022_br0370) 1985
Zhang (10.1016/j.camwa.2025.09.022_br0100) 2017; 74
Achim (10.1016/j.camwa.2025.09.022_br0240) 2001; 20
Lakra (10.1016/j.camwa.2025.09.022_br0340) 2022; 81
Shi (10.1016/j.camwa.2025.09.022_br0500) 2024; 132
Raissi (10.1016/j.camwa.2025.09.022_br0420) 2019; 378
Lee (10.1016/j.camwa.2025.09.022_br0270) 1980; PAMI-2
Guo (10.1016/j.camwa.2025.09.022_br0320) 2015
Prager (10.1016/j.camwa.2025.09.022_br0200) 2002; 415
Jin (10.1016/j.camwa.2025.09.022_br0290) 2000; 4
Karniadakis (10.1016/j.camwa.2025.09.022_br0530) 2021; 3
Gagnon (10.1016/j.camwa.2025.09.022_br0020) 1997
Buades (10.1016/j.camwa.2025.09.022_br0210) 2005
Yuan (10.1016/j.camwa.2025.09.022_br0120) 2023; 61
Perona (10.1016/j.camwa.2025.09.022_br0080) 1990; 12
Kuan (10.1016/j.camwa.2025.09.022_br0260) 1985; 7
Hornik (10.1016/j.camwa.2025.09.022_br0510) 1989; 2
Alparone (10.1016/j.camwa.2025.09.022_br0190) 1998; 19
Witkin (10.1016/j.camwa.2025.09.022_br0070) 1987
Weickert (10.1016/j.camwa.2025.09.022_br0350) 1998
Eslahchi (10.1016/j.camwa.2025.09.022_br0360) 2021; 17
Martin (10.1016/j.camwa.2025.09.022_br0480) 2022; 134
Cai (10.1016/j.camwa.2025.09.022_br0470) 2021; 37
Kharazmi (10.1016/j.camwa.2025.09.022_br0450) 2021; 374
Rudin (10.1016/j.camwa.2025.09.022_br0060) 2003
Yu (10.1016/j.camwa.2025.09.022_br0300) 2002; 11
Namaki (10.1016/j.camwa.2025.09.022_br0490) 2023; 152
Jagtap (10.1016/j.camwa.2025.09.022_br0440) 2020; 365
Yuan (10.1016/j.camwa.2025.09.022_br0140) 2020; 19
Nandal (10.1016/j.camwa.2025.09.022_br0330) 2019; 78
Albisani (10.1016/j.camwa.2025.09.022_br0110) 2025; 190
Deledalle (10.1016/j.camwa.2025.09.022_br0220) 2009; 18
Jagtap (10.1016/j.camwa.2025.09.022_br0430) 2020; 28
References_xml – volume: 7
  start-page: 165
  year: 1985
  end-page: 177
  ident: br0260
  article-title: Adaptive noise smoothing filter for images with signal-dependent noise
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 17
  start-page: 500
  year: 2007
  end-page: 507
  ident: br0010
  article-title: Spatially adaptive wavelet-based method using the Cauchy prior for denoising the sar images
  publication-title: IEEE Trans. Circuits Syst. Video Technol.
– volume: 378
  start-page: 686
  year: 2019
  end-page: 707
  ident: br0420
  article-title: Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
  publication-title: J. Comput. Phys.
– volume: 81
  start-page: 32393
  year: 2022
  end-page: 32420
  ident: br0340
  article-title: Solving a generalized order improved diffusion equation of image denoising using a cenn-based scheme
  publication-title: Multimed. Tools Appl.
– volume: 415
  start-page: 1
  year: 2002
  end-page: 17
  ident: br0200
  article-title: Speckle detection in ultrasound images using first order statistics, GUED/F-INFENG/TR
  publication-title: Intern. Rep.
– volume: 56
  start-page: 3821
  year: 2012
  end-page: 3842
  ident: br0230
  article-title: A new similarity measure for nonlocal filtering in the presence of multiplicative noise
  publication-title: Comput. Stat. Data Anal.
– volume: 66
  start-page: 1145
  year: 1976
  end-page: 1150
  ident: br0160
  article-title: Some fundamental properties of speckle
  publication-title: J. Opt. Soc. Am.
– start-page: 103
  year: 2003
  end-page: 119
  ident: br0060
  article-title: Multiplicative denoising and deblurring: theory and algorithms
  publication-title: Geometric Level Set Methods in Imaging Vision and Graphics
– volume: 1
  start-page: 6
  year: 2013
  end-page: 35
  ident: br0170
  article-title: A tutorial on speckle reduction in synthetic aperture radar images
  publication-title: IEEE Geosci. Remote Sens. Mag.
– volume: 61
  start-page: 1
  year: 2023
  end-page: 16
  ident: br0120
  article-title: Segmentation-guided semantic-aware self-supervised denoising for sar image
  publication-title: IEEE Trans. Geosci. Remote Sens.
– volume: 13
  start-page: 22
  year: 2002
  end-page: 35
  ident: br0400
  article-title: An adaptive finite volume scheme for solving nonlinear diffusion equations in image processing
  publication-title: J. Vis. Commun. Image Represent.
– start-page: 75
  year: 1994
  end-page: 79
  ident: br0030
  article-title: Wavelet based speckle reduction with application to sar based atd/r
  publication-title: 1st International Conference on Image Processing
– volume: PAMI-2
  start-page: 165
  year: 1980
  end-page: 168
  ident: br0270
  article-title: Digital image enhancement and noise filtering by use of local statistics
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 37
  start-page: 1727
  year: 2021
  end-page: 1738
  ident: br0470
  article-title: Physics-informed neural networks (pinns) for fluid mechanics: a review
  publication-title: Acta Mech. Sin.
– volume: 4
  start-page: 157
  year: 1982
  end-page: 166
  ident: br0250
  article-title: A model for radar images and its application to adaptive digital filtering of multiplicative noise
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 190
  start-page: 169
  year: 2025
  end-page: 176
  ident: br0110
  article-title: Self-supervised sar despeckling using deep image prior
  publication-title: Pattern Recognit. Lett.
– volume: 4
  start-page: 298
  year: 2000
  end-page: 305
  ident: br0290
  article-title: An adaptive nonlinear diffusion algorithm for filtering medical images
  publication-title: IEEE Trans. Inf. Technol. Biomed.
– volume: 7
  start-page: 713
  year: 2000
  end-page: 1018
  ident: br0390
  article-title: Finite volume methods
  publication-title: Handb. Numer. Anal.
– volume: 86
  start-page: 2278
  year: 1998
  end-page: 2324
  ident: br0520
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proc. IEEE
– volume: 132
  year: 2024
  ident: br0500
  article-title: Physics-informed convnet: learning physical field from a shallow neural network
  publication-title: Commun. Nonlinear Sci. Numer. Simul.
– volume: 2
  start-page: 359
  year: 1989
  end-page: 366
  ident: br0510
  article-title: Multilayer feedforward networks are universal approximators
  publication-title: Neural Netw.
– volume: 113
  start-page: 315
  year: 2022
  end-page: 331
  ident: br0410
  article-title: Efficient image denoising technique using the meshless method: investigation of operator splitting rbf collocation method for two anisotropic diffusion-based pdes
  publication-title: Comput. Math. Appl.
– start-page: 1
  year: 2024
  end-page: 6
  ident: br0150
  article-title: Unpaired image despeckling based on adversarial speckle generation
  publication-title: IEEE International Conference on Multimedia and Expo
– volume: 18
  start-page: 787
  year: 1999
  end-page: 794
  ident: br0280
  article-title: A novel multiscale nonlinear thresholding method for ultrasonic speckle suppressing
  publication-title: IEEE Trans. Med. Imaging
– volume: 39
  start-page: 62
  year: 2011
  end-page: 74
  ident: br0050
  article-title: A variational model to remove the multiplicative noise in ultrasound images
  publication-title: J. Math. Imaging Vis.
– start-page: 60
  year: 2005
  end-page: 65
  ident: br0210
  article-title: A non-local algorithm for image denoising
  publication-title: IEEE Computer Society Conference on Computer Vision and Pattern Recognition
– volume: 42
  start-page: 130
  year: 1995
  end-page: 132
  ident: br0180
  article-title: Two-dimensional rank-conditioned median filter
  publication-title: IEEE Trans. Circuits Syst. II, Analog Digit. Signal Process.
– volume: 20
  start-page: 772
  year: 2001
  end-page: 783
  ident: br0240
  article-title: Novel Bayesian multiscale method for speckle removal in medical ultrasound images
  publication-title: IEEE Trans. Med. Imaging
– volume: 152
  start-page: 355
  year: 2023
  end-page: 363
  ident: br0490
  article-title: The use of physics-informed neural network approach to image restoration via nonlinear pde tools
  publication-title: Comput. Math. Appl.
– year: 2002
  ident: br0380
  article-title: Finite Element Methods for Partial Differential Equations
– volume: 421
  year: 2024
  ident: br0460
  article-title: Respecting causality for training physics-informed neural networks
  publication-title: Comput. Methods Appl. Mech. Eng.
– volume: 9
  start-page: 1723
  year: 2000
  end-page: 17301
  ident: br0090
  article-title: Fourth-order partial differential equations for noise removal
  publication-title: IEEE Trans. Image Process.
– volume: 78
  start-page: 10717
  year: 2019
  end-page: 10732
  ident: br0330
  article-title: Single image fog removal algorithm in spatial domain using fractional order anisotropic diffusion
  publication-title: Multimed. Tools Appl.
– year: 1985
  ident: br0370
  article-title: Numerical Solution of Partial Differential Equations: Finite Difference Methods
– volume: 11
  start-page: 1260
  year: 2002
  end-page: 1270
  ident: br0300
  article-title: Speckle reducing anisotropic diffusion
  publication-title: IEEE Trans. Image Process.
– volume: 18
  start-page: 2661
  year: 2009
  end-page: 2672
  ident: br0220
  article-title: Iterative weighted maximum likelihood denoising with probabilistic patch-based weights
  publication-title: IEEE Trans. Image Process.
– volume: 19
  start-page: 1
  year: 2020
  end-page: 5
  ident: br0140
  article-title: A practical solution for sar despeckling with adversarial learning generated speckled-to-speckled images
  publication-title: IEEE Geosci. Remote Sens. Lett.
– volume: 134
  start-page: 13
  year: 2022
  ident: br0480
  article-title: Physics-informed neural networks for gravity field modeling of the Earth and Moon
  publication-title: Celest. Mech. Dyn. Astron.
– volume: 365
  year: 2020
  ident: br0440
  article-title: Conservative physics-informed neural networks on discrete domains for conservation laws: applications to forward and inverse problems
  publication-title: Comput. Methods Appl. Mech. Eng.
– volume: 12
  start-page: 629
  year: 1990
  end-page: 639
  ident: br0080
  article-title: Scale-space and edge detection using anisotropic diffusion
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
– volume: 374
  year: 2021
  ident: br0450
  article-title: hp-vpinns: variational physics-informed neural networks with domain decomposition
  publication-title: Comput. Methods Appl. Mech. Eng.
– volume: 218
  start-page: 376
  year: 2024
  end-page: 391
  ident: br0130
  article-title: Contrastive learning for real sar image despeckling
  publication-title: ISPRS J. Photogramm. Remote Sens.
– volume: 28
  start-page: 2002
  year: 2020
  end-page: 2041
  ident: br0430
  article-title: Extended physics-informed neural networks (xpinns): a generalized space-time domain decomposition based deep learning framework for nonlinear partial differential equations
  publication-title: Commun. Comput. Phys.
– year: 1998
  ident: br0350
  article-title: Anisotropic Diffusion in Image Processing
– volume: 43
  start-page: 39
  year: 2012
  end-page: 49
  ident: br0310
  article-title: Adaptive fractional-order multi-scale method for image denoising
  publication-title: J. Math. Imaging Vis.
– volume: 68
  start-page: 925
  year: 2008
  end-page: 946
  ident: br0040
  article-title: A variational approach to removing multiplicative noise
  publication-title: SIAM J. Appl. Math.
– year: 2015
  ident: br0320
  article-title: Fractional Partial Differential Equations and Their Numerical Solutions
– volume: 74
  start-page: 2529
  year: 2017
  end-page: 2545
  ident: br0100
  article-title: An adaptive fourth-order partial differential equation for image denoising
  publication-title: Comput. Math. Appl.
– volume: 17
  start-page: 93
  year: 2021
  end-page: 106
  ident: br0360
  article-title: Application of finite difference method in solving a second- and fourth-order pde blending denoising model
  publication-title: Math. Sci.
– start-page: 329
  year: 1987
  end-page: 332
  ident: br0070
  article-title: Scale-space filtering
  publication-title: Readings in Computer Vision
– volume: 19
  start-page: 89
  year: 1998
  end-page: 96
  ident: br0190
  article-title: Decimated geometric filter for edge-preserving smoothing of non-white image noise
  publication-title: Pattern Recognit. Lett.
– volume: 3
  start-page: 422
  year: 2021
  end-page: 440
  ident: br0530
  article-title: Physics-informed machine learning
  publication-title: Nat. Rev. Phys.
– start-page: 80
  year: 1997
  end-page: 91
  ident: br0020
  article-title: Speckle filtering of sar images: a comparative study between complex-wavelet-based and standard filters
  publication-title: Wavelet Applications in Signal and Image Processing V
– start-page: 103
  year: 2003
  ident: 10.1016/j.camwa.2025.09.022_br0060
  article-title: Multiplicative denoising and deblurring: theory and algorithms
– start-page: 60
  year: 2005
  ident: 10.1016/j.camwa.2025.09.022_br0210
  article-title: A non-local algorithm for image denoising
– volume: PAMI-2
  start-page: 165
  issue: 2
  year: 1980
  ident: 10.1016/j.camwa.2025.09.022_br0270
  article-title: Digital image enhancement and noise filtering by use of local statistics
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.1980.4766994
– volume: 152
  start-page: 355
  year: 2023
  ident: 10.1016/j.camwa.2025.09.022_br0490
  article-title: The use of physics-informed neural network approach to image restoration via nonlinear pde tools
  publication-title: Comput. Math. Appl.
  doi: 10.1016/j.camwa.2023.10.002
– start-page: 80
  year: 1997
  ident: 10.1016/j.camwa.2025.09.022_br0020
  article-title: Speckle filtering of sar images: a comparative study between complex-wavelet-based and standard filters
– volume: 218
  start-page: 376
  year: 2024
  ident: 10.1016/j.camwa.2025.09.022_br0130
  article-title: Contrastive learning for real sar image despeckling
  publication-title: ISPRS J. Photogramm. Remote Sens.
  doi: 10.1016/j.isprsjprs.2024.11.003
– volume: 18
  start-page: 787
  issue: 9
  year: 1999
  ident: 10.1016/j.camwa.2025.09.022_br0280
  article-title: A novel multiscale nonlinear thresholding method for ultrasonic speckle suppressing
  publication-title: IEEE Trans. Med. Imaging
  doi: 10.1109/42.802756
– volume: 2
  start-page: 359
  issue: 5
  year: 1989
  ident: 10.1016/j.camwa.2025.09.022_br0510
  article-title: Multilayer feedforward networks are universal approximators
  publication-title: Neural Netw.
  doi: 10.1016/0893-6080(89)90020-8
– volume: 13
  start-page: 22
  issue: 1–2
  year: 2002
  ident: 10.1016/j.camwa.2025.09.022_br0400
  article-title: An adaptive finite volume scheme for solving nonlinear diffusion equations in image processing
  publication-title: J. Vis. Commun. Image Represent.
  doi: 10.1006/jvci.2001.0502
– year: 2015
  ident: 10.1016/j.camwa.2025.09.022_br0320
– year: 2002
  ident: 10.1016/j.camwa.2025.09.022_br0380
– volume: 19
  start-page: 1
  year: 2020
  ident: 10.1016/j.camwa.2025.09.022_br0140
  article-title: A practical solution for sar despeckling with adversarial learning generated speckled-to-speckled images
  publication-title: IEEE Geosci. Remote Sens. Lett.
– volume: 378
  start-page: 686
  year: 2019
  ident: 10.1016/j.camwa.2025.09.022_br0420
  article-title: Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
  publication-title: J. Comput. Phys.
  doi: 10.1016/j.jcp.2018.10.045
– volume: 86
  start-page: 2278
  issue: 11
  year: 1998
  ident: 10.1016/j.camwa.2025.09.022_br0520
  article-title: Gradient-based learning applied to document recognition
  publication-title: Proc. IEEE
  doi: 10.1109/5.726791
– volume: 42
  start-page: 130
  issue: 2
  year: 1995
  ident: 10.1016/j.camwa.2025.09.022_br0180
  article-title: Two-dimensional rank-conditioned median filter
  publication-title: IEEE Trans. Circuits Syst. II, Analog Digit. Signal Process.
  doi: 10.1109/82.365355
– volume: 365
  year: 2020
  ident: 10.1016/j.camwa.2025.09.022_br0440
  article-title: Conservative physics-informed neural networks on discrete domains for conservation laws: applications to forward and inverse problems
  publication-title: Comput. Methods Appl. Mech. Eng.
  doi: 10.1016/j.cma.2020.113028
– volume: 12
  start-page: 629
  issue: 7
  year: 1990
  ident: 10.1016/j.camwa.2025.09.022_br0080
  article-title: Scale-space and edge detection using anisotropic diffusion
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/34.56205
– volume: 28
  start-page: 2002
  issue: 5
  year: 2020
  ident: 10.1016/j.camwa.2025.09.022_br0430
  article-title: Extended physics-informed neural networks (xpinns): a generalized space-time domain decomposition based deep learning framework for nonlinear partial differential equations
  publication-title: Commun. Comput. Phys.
  doi: 10.4208/cicp.OA-2020-0164
– start-page: 75
  year: 1994
  ident: 10.1016/j.camwa.2025.09.022_br0030
  article-title: Wavelet based speckle reduction with application to sar based atd/r
– year: 1985
  ident: 10.1016/j.camwa.2025.09.022_br0370
– volume: 19
  start-page: 89
  issue: 1
  year: 1998
  ident: 10.1016/j.camwa.2025.09.022_br0190
  article-title: Decimated geometric filter for edge-preserving smoothing of non-white image noise
  publication-title: Pattern Recognit. Lett.
  doi: 10.1016/S0167-8655(97)00157-8
– volume: 113
  start-page: 315
  year: 2022
  ident: 10.1016/j.camwa.2025.09.022_br0410
  article-title: Efficient image denoising technique using the meshless method: investigation of operator splitting rbf collocation method for two anisotropic diffusion-based pdes
  publication-title: Comput. Math. Appl.
  doi: 10.1016/j.camwa.2022.03.013
– volume: 4
  start-page: 298
  issue: 4
  year: 2000
  ident: 10.1016/j.camwa.2025.09.022_br0290
  article-title: An adaptive nonlinear diffusion algorithm for filtering medical images
  publication-title: IEEE Trans. Inf. Technol. Biomed.
  doi: 10.1109/4233.897062
– volume: 66
  start-page: 1145
  issue: 11
  year: 1976
  ident: 10.1016/j.camwa.2025.09.022_br0160
  article-title: Some fundamental properties of speckle
  publication-title: J. Opt. Soc. Am.
  doi: 10.1364/JOSA.66.001145
– volume: 43
  start-page: 39
  year: 2012
  ident: 10.1016/j.camwa.2025.09.022_br0310
  article-title: Adaptive fractional-order multi-scale method for image denoising
  publication-title: J. Math. Imaging Vis.
  doi: 10.1007/s10851-011-0285-z
– volume: 37
  start-page: 1727
  issue: 12
  year: 2021
  ident: 10.1016/j.camwa.2025.09.022_br0470
  article-title: Physics-informed neural networks (pinns) for fluid mechanics: a review
  publication-title: Acta Mech. Sin.
  doi: 10.1007/s10409-021-01148-1
– start-page: 329
  year: 1987
  ident: 10.1016/j.camwa.2025.09.022_br0070
  article-title: Scale-space filtering
– volume: 61
  start-page: 1
  year: 2023
  ident: 10.1016/j.camwa.2025.09.022_br0120
  article-title: Segmentation-guided semantic-aware self-supervised denoising for sar image
  publication-title: IEEE Trans. Geosci. Remote Sens.
– volume: 20
  start-page: 772
  issue: 8
  year: 2001
  ident: 10.1016/j.camwa.2025.09.022_br0240
  article-title: Novel Bayesian multiscale method for speckle removal in medical ultrasound images
  publication-title: IEEE Trans. Med. Imaging
  doi: 10.1109/42.938245
– volume: 74
  start-page: 2529
  issue: 10
  year: 2017
  ident: 10.1016/j.camwa.2025.09.022_br0100
  article-title: An adaptive fourth-order partial differential equation for image denoising
  publication-title: Comput. Math. Appl.
  doi: 10.1016/j.camwa.2017.07.036
– volume: 374
  year: 2021
  ident: 10.1016/j.camwa.2025.09.022_br0450
  article-title: hp-vpinns: variational physics-informed neural networks with domain decomposition
  publication-title: Comput. Methods Appl. Mech. Eng.
  doi: 10.1016/j.cma.2020.113547
– start-page: 1
  year: 2024
  ident: 10.1016/j.camwa.2025.09.022_br0150
  article-title: Unpaired image despeckling based on adversarial speckle generation
– volume: 11
  start-page: 1260
  issue: 11
  year: 2002
  ident: 10.1016/j.camwa.2025.09.022_br0300
  article-title: Speckle reducing anisotropic diffusion
  publication-title: IEEE Trans. Image Process.
  doi: 10.1109/TIP.2002.804276
– volume: 415
  start-page: 1
  year: 2002
  ident: 10.1016/j.camwa.2025.09.022_br0200
  article-title: Speckle detection in ultrasound images using first order statistics, GUED/F-INFENG/TR
  publication-title: Intern. Rep.
– volume: 56
  start-page: 3821
  issue: 12
  year: 2012
  ident: 10.1016/j.camwa.2025.09.022_br0230
  article-title: A new similarity measure for nonlocal filtering in the presence of multiplicative noise
  publication-title: Comput. Stat. Data Anal.
  doi: 10.1016/j.csda.2012.05.009
– volume: 1
  start-page: 6
  issue: 3
  year: 2013
  ident: 10.1016/j.camwa.2025.09.022_br0170
  article-title: A tutorial on speckle reduction in synthetic aperture radar images
  publication-title: IEEE Geosci. Remote Sens. Mag.
  doi: 10.1109/MGRS.2013.2277512
– volume: 421
  year: 2024
  ident: 10.1016/j.camwa.2025.09.022_br0460
  article-title: Respecting causality for training physics-informed neural networks
  publication-title: Comput. Methods Appl. Mech. Eng.
  doi: 10.1016/j.cma.2024.116813
– volume: 4
  start-page: 157
  issue: 2
  year: 1982
  ident: 10.1016/j.camwa.2025.09.022_br0250
  article-title: A model for radar images and its application to adaptive digital filtering of multiplicative noise
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.1982.4767223
– volume: 134
  start-page: 13
  issue: 2
  year: 2022
  ident: 10.1016/j.camwa.2025.09.022_br0480
  article-title: Physics-informed neural networks for gravity field modeling of the Earth and Moon
  publication-title: Celest. Mech. Dyn. Astron.
  doi: 10.1007/s10569-022-10069-5
– volume: 18
  start-page: 2661
  issue: 12
  year: 2009
  ident: 10.1016/j.camwa.2025.09.022_br0220
  article-title: Iterative weighted maximum likelihood denoising with probabilistic patch-based weights
  publication-title: IEEE Trans. Image Process.
  doi: 10.1109/TIP.2009.2029593
– volume: 68
  start-page: 925
  issue: 4
  year: 2008
  ident: 10.1016/j.camwa.2025.09.022_br0040
  article-title: A variational approach to removing multiplicative noise
  publication-title: SIAM J. Appl. Math.
  doi: 10.1137/060671814
– volume: 7
  start-page: 165
  issue: 2
  year: 1985
  ident: 10.1016/j.camwa.2025.09.022_br0260
  article-title: Adaptive noise smoothing filter for images with signal-dependent noise
  publication-title: IEEE Trans. Pattern Anal. Mach. Intell.
  doi: 10.1109/TPAMI.1985.4767641
– volume: 190
  start-page: 169
  year: 2025
  ident: 10.1016/j.camwa.2025.09.022_br0110
  article-title: Self-supervised sar despeckling using deep image prior
  publication-title: Pattern Recognit. Lett.
  doi: 10.1016/j.patrec.2025.02.021
– volume: 17
  start-page: 93
  year: 2021
  ident: 10.1016/j.camwa.2025.09.022_br0360
  article-title: Application of finite difference method in solving a second- and fourth-order pde blending denoising model
  publication-title: Math. Sci.
  doi: 10.1007/s40096-021-00444-y
– volume: 7
  start-page: 713
  year: 2000
  ident: 10.1016/j.camwa.2025.09.022_br0390
  article-title: Finite volume methods
  publication-title: Handb. Numer. Anal.
– volume: 9
  start-page: 1723
  issue: 10
  year: 2000
  ident: 10.1016/j.camwa.2025.09.022_br0090
  article-title: Fourth-order partial differential equations for noise removal
  publication-title: IEEE Trans. Image Process.
  doi: 10.1109/83.869184
– year: 1998
  ident: 10.1016/j.camwa.2025.09.022_br0350
– volume: 17
  start-page: 500
  issue: 4
  year: 2007
  ident: 10.1016/j.camwa.2025.09.022_br0010
  article-title: Spatially adaptive wavelet-based method using the Cauchy prior for denoising the sar images
  publication-title: IEEE Trans. Circuits Syst. Video Technol.
  doi: 10.1109/TCSVT.2006.888020
– volume: 81
  start-page: 32393
  issue: 22
  year: 2022
  ident: 10.1016/j.camwa.2025.09.022_br0340
  article-title: Solving a generalized order improved diffusion equation of image denoising using a cenn-based scheme
  publication-title: Multimed. Tools Appl.
  doi: 10.1007/s11042-022-12998-7
– volume: 3
  start-page: 422
  issue: 6
  year: 2021
  ident: 10.1016/j.camwa.2025.09.022_br0530
  article-title: Physics-informed machine learning
  publication-title: Nat. Rev. Phys.
  doi: 10.1038/s42254-021-00314-5
– volume: 39
  start-page: 62
  issue: 1
  year: 2011
  ident: 10.1016/j.camwa.2025.09.022_br0050
  article-title: A variational model to remove the multiplicative noise in ultrasound images
  publication-title: J. Math. Imaging Vis.
  doi: 10.1007/s10851-010-0225-3
– volume: 132
  year: 2024
  ident: 10.1016/j.camwa.2025.09.022_br0500
  article-title: Physics-informed convnet: learning physical field from a shallow neural network
  publication-title: Commun. Nonlinear Sci. Numer. Simul.
  doi: 10.1016/j.cnsns.2024.107911
– volume: 78
  start-page: 10717
  year: 2019
  ident: 10.1016/j.camwa.2025.09.022_br0330
  article-title: Single image fog removal algorithm in spatial domain using fractional order anisotropic diffusion
  publication-title: Multimed. Tools Appl.
  doi: 10.1007/s11042-018-6576-2
SSID ssj0004320
Score 2.4596364
Snippet In this study, we propose a new data-driven algorithm for the Perona-Malik image despeckling problem. The advantage of the proposed algorithm over neural...
SourceID crossref
elsevier
SourceType Index Database
Publisher
StartPage 202
SubjectTerms Convolutional neural networks
Diffusion
Finite difference method
Image despeckling
PDEs
Physics-informed machine learning
Title A data-driven algorithm for solving image despeckling PDE model using physics-informed ConvNet
URI https://dx.doi.org/10.1016/j.camwa.2025.09.022
Volume 200
WOSCitedRecordID wos001585693000002&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: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  issn: 0898-1221
  databaseCode: AIEXJ
  dateStart: 20211209
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.sciencedirect.com
  omitProxy: false
  ssIdentifier: ssj0004320
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3Nb9MwFLdKx4EL34jxJR-4FUuJnTT2sRpFA6FqEgP1RGTHNlu3plPblV32v_Nsx2m2oYkduERpojof76fn93t572eE3ttc2tSmnMiC5ySThhPFWEbyCqb71BSp9Mt0_vhaTCZ8OhUHvd5l7IXZnBZ1zS8uxNl_NTUcA2O71tk7mLsdFA7APhgdtmB22P6T4UcDV_VJ9NL5sYE8_bUA_n809_WEcGWfQDieu1IdbVyf5YlvSD_4OA6r4gzOffYgZDxWJAirujTwot5MzJVMflwRYuXxM28FYGPHXOfbuNeDupp4aIu794Gua-DXR1ugtue-yXpmzKabm6C5q_MI3ZnRhQngqDT0QEd_G6RJW4-Z0M7kS4NQwA2_HlIMM-Ds899OLYrmXp2W0u00Fj_dX5vd2prDWM42K_0gpRukTEQJg9xDO7TIBe-jndHn8fTLtrGWBV3P-BhRtsoXCN64l7-HNp1w5fAxetjwDDwK-HiCeqZ-ih5Fi-HGpT9DP0e4AxfcwgWD2XEDF-zhgjtwwQAX7OGCPVzwdbjgBi7P0fdP48O9fdKsuUEqIJdrIm1SSZUrSw23QkjYcZRgaGhlKiqAj3GmIarVBVWpMQnTilvFjYWfhbaMvUD9elGblwhLlsmMSSE055kdWpUXyVABm8uYSXSqdtGH-LrKsyCtUt5ipF00jK-0bKLDEPWVAJLb_vjqbtd5jR5sofwG9dfLc_MW3a826-PV8l2DkD8v1oj0
linkProvider Elsevier
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=A+data-driven+algorithm+for+solving+image+despeckling+PDE+model+using+physics-informed+ConvNet&rft.jtitle=Computers+%26+mathematics+with+applications+%281987%29&rft.au=Kumar%2C+Haridarshan&rft.au=Kumar%2C+Sanjeev&rft.date=2025-12-15&rft.issn=0898-1221&rft.volume=200&rft.spage=202&rft.epage=227&rft_id=info:doi/10.1016%2Fj.camwa.2025.09.022&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_camwa_2025_09_022
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0898-1221&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0898-1221&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0898-1221&client=summon