SpaceNet 9—Cross-Sensor Alignment of Optical and SAR Imagery

Ronny Hansch; Jacob Arndt; Abhishek Potnis; Philipe Dias; Peter Novotny; Fabio Pacifici; Todd M. Bacastow

doi:10.1109/JSTARS.2026.3666103

SpaceNet 9—Cross-Sensor Alignment of Optical and SAR Imagery

Ronny Hansch
, Jacob Arndt
, Abhishek Potnis
, Philipe Dias
, Peter Novotny
, Fabio Pacifici
, Todd M. Bacastow

GeoAI

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

Abstract

Precise registration of high-resolution synthetic aperture radar (SAR) and optical imagery is necessary for realizing the full potential and benefits of multimodal image analysis. However, two significant challenges presently exist. First, there is a lack of annotated datasets and benchmarks available for high-resolution SAR–optical image registration. Second, an assessment of efficient and reliable image registration methods that can precisely align these modalities is lacking. Here, we present a holistic description of the SpaceNet 9 Challenge and its results. We present a description of the dataset and baseline algorithm along with the results of the challenge, including a description of the winning algorithms. We release the SpaceNet 9 dataset along with open-sourcing the winning algorithms and baseline. The objective of SpaceNet 9 was to compute a dense displacement map that indicates the shift needed to align pixels in an optical image to the pixels in a SAR image. The challenge launched in April 2025 and was active for approximately two months. The top five solutions reduced image alignment error from approximately 34 m to under 13 m for public and private test data, with the best results obtaining a registration error of only 8.5 and 6.7 m on the public testing and private testing dataset, respectively. Usage of pretrained image matching models, robust outlier rejection with RANSAC, and estimating local displacement were common among the top solutions. The results of this challenge provide insight into high-resolution SAR–optical image registration and offer opportunities for future benchmarking in this domain. The baseline algorithm, winning solutions, and datasets are available at https://spacenet.ai/sn9-challenge/.

Original language	English
Pages (from-to)	11491-11502
Number of pages	12
Journal	IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume	19
DOIs	https://doi.org/10.1109/JSTARS.2026.3666103
State	Published - 2026

Funding

We acknowledge that this manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. We acknowledge the contributions of all of the SpaceNet partners including AWS, IEEE GRSS, Oak Ridge National Laboratory, OGC, Topcoder, UMBRA, and Vantor. We acknowledge that this manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan). Received 20 November 2025; revised 2 February 2026; accepted 13 February 2026. Date of publication 18 February 2026; date of current version 9 April 2026. We acknowledge that this manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. (Corresponding author: Ronny Hänsch.) Ronny Hänsch is with German Aerospace Center (DLR), 82234 Oberpfaffenhofen-Weßling, Germany (e-mail: [email protected]).

Keywords

Benchmark datasets
cross-modal
image registration
multimodal learning
optical
synthetic aperture radar (SAR)

Access to Document

10.1109/JSTARS.2026.3666103

Cite this

@article{269146f839a2431fb5e12671641a0a33,

title = "SpaceNet 9—Cross-Sensor Alignment of Optical and SAR Imagery",

abstract = "Precise registration of high-resolution synthetic aperture radar (SAR) and optical imagery is necessary for realizing the full potential and benefits of multimodal image analysis. However, two significant challenges presently exist. First, there is a lack of annotated datasets and benchmarks available for high-resolution SAR–optical image registration. Second, an assessment of efficient and reliable image registration methods that can precisely align these modalities is lacking. Here, we present a holistic description of the SpaceNet 9 Challenge and its results. We present a description of the dataset and baseline algorithm along with the results of the challenge, including a description of the winning algorithms. We release the SpaceNet 9 dataset along with open-sourcing the winning algorithms and baseline. The objective of SpaceNet 9 was to compute a dense displacement map that indicates the shift needed to align pixels in an optical image to the pixels in a SAR image. The challenge launched in April 2025 and was active for approximately two months. The top five solutions reduced image alignment error from approximately 34 m to under 13 m for public and private test data, with the best results obtaining a registration error of only 8.5 and 6.7 m on the public testing and private testing dataset, respectively. Usage of pretrained image matching models, robust outlier rejection with RANSAC, and estimating local displacement were common among the top solutions. The results of this challenge provide insight into high-resolution SAR–optical image registration and offer opportunities for future benchmarking in this domain. The baseline algorithm, winning solutions, and datasets are available at https://spacenet.ai/sn9-challenge/.",

keywords = "Benchmark datasets, cross-modal, image registration, multimodal learning, optical, synthetic aperture radar (SAR)",

author = "Ronny Hansch and Jacob Arndt and Abhishek Potnis and Philipe Dias and Peter Novotny and Fabio Pacifici and Bacastow, \{Todd M.\}",

note = "Publisher Copyright: {\textcopyright} 2008-2012 IEEE.",

year = "2026",

doi = "10.1109/JSTARS.2026.3666103",

language = "English",

volume = "19",

pages = "11491--11502",

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

issn = "1939-1404",

publisher = "Institute of Electrical and Electronics Engineers",

}

TY - JOUR

T1 - SpaceNet 9—Cross-Sensor Alignment of Optical and SAR Imagery

AU - Hansch, Ronny

AU - Arndt, Jacob

AU - Potnis, Abhishek

AU - Dias, Philipe

AU - Novotny, Peter

AU - Pacifici, Fabio

AU - Bacastow, Todd M.

PY - 2026

Y1 - 2026

N2 - Precise registration of high-resolution synthetic aperture radar (SAR) and optical imagery is necessary for realizing the full potential and benefits of multimodal image analysis. However, two significant challenges presently exist. First, there is a lack of annotated datasets and benchmarks available for high-resolution SAR–optical image registration. Second, an assessment of efficient and reliable image registration methods that can precisely align these modalities is lacking. Here, we present a holistic description of the SpaceNet 9 Challenge and its results. We present a description of the dataset and baseline algorithm along with the results of the challenge, including a description of the winning algorithms. We release the SpaceNet 9 dataset along with open-sourcing the winning algorithms and baseline. The objective of SpaceNet 9 was to compute a dense displacement map that indicates the shift needed to align pixels in an optical image to the pixels in a SAR image. The challenge launched in April 2025 and was active for approximately two months. The top five solutions reduced image alignment error from approximately 34 m to under 13 m for public and private test data, with the best results obtaining a registration error of only 8.5 and 6.7 m on the public testing and private testing dataset, respectively. Usage of pretrained image matching models, robust outlier rejection with RANSAC, and estimating local displacement were common among the top solutions. The results of this challenge provide insight into high-resolution SAR–optical image registration and offer opportunities for future benchmarking in this domain. The baseline algorithm, winning solutions, and datasets are available at https://spacenet.ai/sn9-challenge/.

AB - Precise registration of high-resolution synthetic aperture radar (SAR) and optical imagery is necessary for realizing the full potential and benefits of multimodal image analysis. However, two significant challenges presently exist. First, there is a lack of annotated datasets and benchmarks available for high-resolution SAR–optical image registration. Second, an assessment of efficient and reliable image registration methods that can precisely align these modalities is lacking. Here, we present a holistic description of the SpaceNet 9 Challenge and its results. We present a description of the dataset and baseline algorithm along with the results of the challenge, including a description of the winning algorithms. We release the SpaceNet 9 dataset along with open-sourcing the winning algorithms and baseline. The objective of SpaceNet 9 was to compute a dense displacement map that indicates the shift needed to align pixels in an optical image to the pixels in a SAR image. The challenge launched in April 2025 and was active for approximately two months. The top five solutions reduced image alignment error from approximately 34 m to under 13 m for public and private test data, with the best results obtaining a registration error of only 8.5 and 6.7 m on the public testing and private testing dataset, respectively. Usage of pretrained image matching models, robust outlier rejection with RANSAC, and estimating local displacement were common among the top solutions. The results of this challenge provide insight into high-resolution SAR–optical image registration and offer opportunities for future benchmarking in this domain. The baseline algorithm, winning solutions, and datasets are available at https://spacenet.ai/sn9-challenge/.

KW - Benchmark datasets

KW - cross-modal

KW - image registration

KW - multimodal learning

KW - optical

KW - synthetic aperture radar (SAR)

UR - https://www.scopus.com/pages/publications/105030691463

U2 - 10.1109/JSTARS.2026.3666103

DO - 10.1109/JSTARS.2026.3666103

M3 - Article

AN - SCOPUS:105030691463

SN - 1939-1404

VL - 19

SP - 11491

EP - 11502

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 -

SpaceNet 9—Cross-Sensor Alignment of Optical and SAR Imagery

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this