Fixed-Quality Compression of Remote Sensing Images With Neural Networks

Sebastià Mijares Verdú; Marie Chabert; Thomas Oberlin; Joan Serra Sagrista

doi:10.1109/JSTARS.2024.3422215

Fixed-Quality Compression of Remote Sensing Images With Neural Networks

Sebastià Mijares Verdú^*, Marie Chabert, Thomas Oberlin, Joan Serra Sagrista

^*Corresponding author for this work

Department of Information and Communications Engineering

Research output: Contribution to journal › Article › Research › peer-review

1 Citation (Scopus)

Abstract

Fixed-quality image compression is a coding paradigm where the tolerated introduced distortion is set by the user. This article proposes a novel fixed-quality compression method for remote sensing images. It is based on a neural architecture we have recently proposed for multirate satellite image compression. In this article, we show how to efficiently estimate the reconstruction quality using an appropriate statistical model. The performance of our approach is assessed and compared against recent fixed-quality coding techniques and standards in terms of accuracy and rate-distortion, as well as with recent machine learning compression methods in rate-distortion, showing competitive results. In particular, the proposed method does not introduce artifacts even when coding neighboring areas at different qualities.

Original language	English
Pages (from-to)	12169-12180
Number of pages	12
Journal	IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume	17
DOIs	https://doi.org/10.1109/JSTARS.2024.3422215
Publication status	Published - 3 Jul 2024

Keywords

Image coding
Remote sensing
Standards
Neural networks
Codecs
Image reconstruction
Distortion
Neural network applications
Data compression
Optical data processing

Access to Document

10.1109/JSTARS.2024.3422215Licence: CC BY

COTS AND LEARNING IN NANOSATELLITE DATA COMPRESSION
Serra Sagrista, J. (Principal Investigator), Blanes Garcia, I. (Principal Investigator 2), Auli Llinas, F. (Investigator), Bartrina Rapesta, J. (Investigator), Blasco Andreo, N. (Collaborator), Capella del Solar, A. (Collaborator), Cantón Simó, S. (Collaborator), Chow , H. F. K. (Collaborator), Hernández Cabronero, M. (Collaborator), Jara Lorente, E. (Collaborator), Gururajan ., A. K. (Collaborator), Llabrés Brustenga, A. (Collaborator), Maireles Gonzalez, O. (Collaborator), Mijares Verdú, S. (Collaborator), Ribas Nicolau, R. (Collaborator), Tzamarias , D. E. O. (Collaborator), Sánchez Silva, V. F. (Collaborator), Martinez Torres, M. I. (Collaborator), Marcellin, M. W. (Collaborator), Hernández Cabronero, M. (Investigator), Alegre Alarza, D. (Collaborator), Fernandez Mellado, X. (Collaborator) & Toro Valdivia, M. C. D. (Collaborator)
1/09/22 → 31/08/25
Project: Research Projects and Other Grants

Remote sensing data compression with neural networks
Mijares i Verdú, S. (Author), Serra Sagrista, J. (Director) & Bartrina Rapesta, J. (Director), 25 Jul 2024
Student thesis: Doctoral thesis

File

Cite this

@article{e76d667226f44744aff97c6d1c3beeff,

title = "Fixed-Quality Compression of Remote Sensing Images With Neural Networks",

abstract = "Fixed-quality image compression is a coding paradigm where the tolerated introduced distortion is set by the user. This article proposes a novel fixed-quality compression method for remote sensing images. It is based on a neural architecture we have recently proposed for multirate satellite image compression. In this article, we show how to efficiently estimate the reconstruction quality using an appropriate statistical model. The performance of our approach is assessed and compared against recent fixed-quality coding techniques and standards in terms of accuracy and rate-distortion, as well as with recent machine learning compression methods in rate-distortion, showing competitive results. In particular, the proposed method does not introduce artifacts even when coding neighboring areas at different qualities.",

keywords = "Image coding, Remote sensing, Standards, Neural networks, Codecs, Image reconstruction, Distortion, Neural network applications, Data compression, Optical data processing",

author = "{Mijares Verd{\'u}}, Sebasti{\`a} and Marie Chabert and Thomas Oberlin and {Serra Sagrista}, Joan",

note = "DBLP License: DBLP's bibliographic metadata records provided through http://dblp.org/ are distributed under a Creative Commons CC0 1.0 Universal Public Domain Dedication. Although the bibliographic metadata records are provided consistent with CC0 1.0 Dedication, the content described by the metadata records is not. Content may be subject to copyright, rights of privacy, rights of publicity and other restrictions.",

year = "2024",

month = jul,

day = "3",

doi = "10.1109/JSTARS.2024.3422215",

language = "English",

volume = "17",

pages = "12169--12180",

journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",

issn = "1939-1404",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Fixed-Quality Compression of Remote Sensing Images With Neural Networks

AU - Mijares Verdú, Sebastià

AU - Chabert, Marie

AU - Oberlin, Thomas

AU - Serra Sagrista, Joan

N1 - DBLP License: DBLP's bibliographic metadata records provided through http://dblp.org/ are distributed under a Creative Commons CC0 1.0 Universal Public Domain Dedication. Although the bibliographic metadata records are provided consistent with CC0 1.0 Dedication, the content described by the metadata records is not. Content may be subject to copyright, rights of privacy, rights of publicity and other restrictions.

PY - 2024/7/3

Y1 - 2024/7/3

N2 - Fixed-quality image compression is a coding paradigm where the tolerated introduced distortion is set by the user. This article proposes a novel fixed-quality compression method for remote sensing images. It is based on a neural architecture we have recently proposed for multirate satellite image compression. In this article, we show how to efficiently estimate the reconstruction quality using an appropriate statistical model. The performance of our approach is assessed and compared against recent fixed-quality coding techniques and standards in terms of accuracy and rate-distortion, as well as with recent machine learning compression methods in rate-distortion, showing competitive results. In particular, the proposed method does not introduce artifacts even when coding neighboring areas at different qualities.

AB - Fixed-quality image compression is a coding paradigm where the tolerated introduced distortion is set by the user. This article proposes a novel fixed-quality compression method for remote sensing images. It is based on a neural architecture we have recently proposed for multirate satellite image compression. In this article, we show how to efficiently estimate the reconstruction quality using an appropriate statistical model. The performance of our approach is assessed and compared against recent fixed-quality coding techniques and standards in terms of accuracy and rate-distortion, as well as with recent machine learning compression methods in rate-distortion, showing competitive results. In particular, the proposed method does not introduce artifacts even when coding neighboring areas at different qualities.

KW - Image coding

KW - Remote sensing

KW - Standards

KW - Neural networks

KW - Codecs

KW - Image reconstruction

KW - Distortion

KW - Neural network applications

KW - Data compression

KW - Optical data processing

U2 - 10.1109/JSTARS.2024.3422215

DO - 10.1109/JSTARS.2024.3422215

M3 - Article

SN - 1939-1404

VL - 17

SP - 12169

EP - 12180

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

ER -

Fixed-Quality Compression of Remote Sensing Images With Neural Networks

Abstract

Keywords

Access to Document

Fingerprint

Projects

COTS AND LEARNING IN NANOSATELLITE DATA COMPRESSION

Student theses

Remote sensing data compression with neural networks

Cite this