Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Review of compression approaches for optical remote-sensing images from post-processing perspective

Chen, Xinyi ; Zhou, Han ; Li, Yong LU ; Wang, Zhengjun ; Shan, Fan ; Lin, Weifeng ; Da, Hongju and Li, Jin (2026) In Geocarto International 41(1).
Abstract

Image compression is crucial for storing remote sensing data in space-constrained on-orbit systems. Current methods primarily follow two directions. The first combines wavelet transforms with post-processing, leveraging wavelets' sparse representation for multi-resolution analysis and efficient hardware implementation. The second employs compressed sensing (CS), which shifts complexity to the decoder; its integration with deep learning has advanced the field. Deep learning itself is emerging as a third framework, using joint encoding-decoding to overcome the limitations of fixed transform bases. This paper outlines these methods' principles, configurations and developments.

Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
compression, Image, remote sensing
in
Geocarto International
volume
41
issue
1
article number
2647547
publisher
Taylor & Francis
external identifiers
  • scopus:105033949198
ISSN
1010-6049
DOI
10.1080/10106049.2026.2647547
language
English
LU publication?
yes
additional info
Publisher Copyright: © 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
id
1328580f-604f-4114-ab34-44ccb60b9ea8
date added to LUP
2026-07-01 12:45:16
date last changed
2026-07-01 15:12:12
@article{1328580f-604f-4114-ab34-44ccb60b9ea8,
  abstract     = {{<p>Image compression is crucial for storing remote sensing data in space-constrained on-orbit systems. Current methods primarily follow two directions. The first combines wavelet transforms with post-processing, leveraging wavelets' sparse representation for multi-resolution analysis and efficient hardware implementation. The second employs compressed sensing (CS), which shifts complexity to the decoder; its integration with deep learning has advanced the field. Deep learning itself is emerging as a third framework, using joint encoding-decoding to overcome the limitations of fixed transform bases. This paper outlines these methods' principles, configurations and developments.</p>}},
  author       = {{Chen, Xinyi and Zhou, Han and Li, Yong and Wang, Zhengjun and Shan, Fan and Lin, Weifeng and Da, Hongju and Li, Jin}},
  issn         = {{1010-6049}},
  keywords     = {{compression; Image; remote sensing}},
  language     = {{eng}},
  number       = {{1}},
  publisher    = {{Taylor & Francis}},
  series       = {{Geocarto International}},
  title        = {{Review of compression approaches for optical remote-sensing images from post-processing perspective}},
  url          = {{http://dx.doi.org/10.1080/10106049.2026.2647547}},
  doi          = {{10.1080/10106049.2026.2647547}},
  volume       = {{41}},
  year         = {{2026}},
}