Evolution of AI-based change detection for low-to-medium resolution satellite imagery and design strategies for user-customized generalized platforms

Heein Yang; Daehee Kim

doi:10.23386/joss.2026.3.1.002

All Issue

2026 Vol.3, Issue 1 Preview Page Next Page

Evolution of AI-based change detection for low-to-medium resolution satellite imagery and design strategies for user-customized generalized platforms

30 June 2026. pp. 13-21

PDF XML

Abstract

With the rapid advancement of Earth observation satellite technology and the widespread operation of small satellite constellations, optical image data capable of monitoring wide areas on a regular and continuous basis is exploding. The accumulation of this time-series data has highlighted the importance of change detection technology, which identifies and interprets changes in the Earth’s surface in a rapidly changing global environment driven by climate change, large-scale natural disasters, urbanization, and international conflict. Change detection based on low- to mid-resolution optical satellite imagery offers significant value for wide-area and long-term monitoring. However, it presents technical challenges that differ from high-resolution change detection, which focuses on detecting the presence and changes of targets, due to differences in pixel characteristics across time periods, location alignment errors, and radiometric mismatches caused by the atmosphere, clouds, and seasonality. This paper assesses the current state of change detection technology based on low- to mid-resolution satellite imagery and provides an in-depth analysis of the evolution of deep learning architectures. This study quantitatively and qualitatively examines how the evolution of deep learning models has contributed to improved change detection performance, starting with traditional statistical techniques, moving on to Convolutional Neural Networks (CNNs), UNet-based Siamese networks, and the Vision Transformer (ViT), which captures global context. Finally, we present the State Space Model (SSM)-based Mamba architecture, which has recently become a leader in change detection, demonstrating superior spatial feature capture while maintaining linear computational complexity. However, despite the advancements in deep learning model architecture, a new model-centric approach alone cannot fully satisfy the complex and diverse requirements of real-world applications. This study emphasizes that while improving the performance of the AI model itself is crucial, designing a monitoring system tailored to the end user’s needs (analysis objectives, target objects, and resolution) is a critical factor in determining practical success or failure. Furthermore, the paper concludes by suggesting the necessity of building a ‘next-generation generalized change detection platform’ that will accommodate various foundation models and heterogeneous sensor data (hyperspectral, SAR, etc.) and provide general-purpose functions by integrating intelligent preprocessing such as cloud and haze removal, and its specific development direction.

Keywords

Satellite change detection

Multiresolution imagery

Siamese networks

Transformers

Semi-supervised learning

Imbalance

Super-resolution

Attention

Deformable convolution

References

H. Chen, and Z. Shi, A spatial-temporal attention-based method and a new dataset for remote sensing image change detection. Remote Sensing. 12, 2020, Article 1662. DOI: 10.3390/rs12101662.

10.3390/rs12101662

S. Ji, S. Wei, and M. Lu, Fully convolutional networks for multisource building extraction from an open aerial and satellite imagery dataset. IEEE Transactions on Geoscience and Remote Sensing. 57, 2019, 574-586. DOI: 10.1109/ TGRS.2018.2858817

10.1109/TGRS.2018.2858817

L. Shen, Y. Lu, H. Chen, H. Wei, D. Xie, Y. Yue, R. Chen, S. Lv, and B. Jiang, S2Looking: A satellite side-looking dataset for building change detection. Remote Sensing. 13(24), 2021, Article 5094. DOI: 10.3390/rs13245094.

10.3390/rs13245094

L. Shen, Y. Lu, H. Chen, H. Wei, D. Xie, J. Yue, R. Chen, S. Lv, and B. Jiang, Building change detection for remote sensing images using a dual-task constrained deep Siamese convolutional network model. IEEE Geoscience and Remote Sensing Letters. 17(5), 2021, pp. 811-815.

10.1109/LGRS.2020.2988032

J. Long, E. Shelhamer, and T. Darrell, Fully convolutional networks for semantic segmentation, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 7-12 June, 2015, pp. 3431-3440. DOI: 10.1109/CVPR.2015.7298965.

10.1109/CVPR.2015.7298965

R.C. Daudt, B. Le Saux, and A. Boulch, Fully convolutional siamese networks for change detection, 2018 IEEE International Conference on Image Processing (ICIP), 7-10 October, 2018, pp. 4063-4067. DOI: 10.1109/ICIP.2018.8451652.

10.1109/ICIP.2018.8451652

R. C. Daudt, B. Le Saux, A. Boulch, and Y. Gousseau, Urban change detection for multispectral earth observation using convolutional neural networks, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 22-27 July, 2018, pp. 2115-2118. DOI: 10.1109/IGARSS.2018.8518015.

10.1109/IGARSS.2018.8518015

C. Benedek and T. Szirányi, Change detection in optical aerial images by a multilayer conditional mixed Markov model. IEEE Transactions on Geoscience and Remote Sensing. 47, 2009, pp. 3416-3430. DOI: 10.1109/TGRS.2009.2022633.

10.1109/TGRS.2009.2022633

A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, Kaiser, and I. Polosukhin, Attention is all you need. Advances in neural information processing systems. 30, 2017, Article 30.

H. Chen and Z. Shi, Remote sensing image change detection with transformers. IEEE Transactions on Geoscience and Remote Sensing. 60, 2021, pp. 1-14. DOI: 10.1109/TGRS.2021.3095166.

10.1109/TGRS.2021.3095166

A. Gu and T. Dao, Mamba: Linear-time sequence modeling with selective state spaces. 2023. arXiv: 2312.00752.

Y. Liu, Y. Tian, Y. Zhao, H. Yu, L. Xie, Y. Wang, Q. Ye, J. Jiao, and Y. Liu, VMamba: Visual state space model. 2024. arXiv: 2401.10166.

10.52202/079017-3273

H. Chen, J. Song, C. Han, J. Xia, and N. Yokoya, ChangeMamba: Remote Sensing Change Detection With Spatiotemporal State Space Model. IEEE Transactions on Geoscience and Remote Sensing. 62, 2024, pp. 1-20

10.1109/TGRS.2024.3417253

Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin, and B. Guo, Swin Transformer: Hierarchical vision transformer using shifted windows, 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 10-17, October, 2021, pp. 10012-10022.

10.1109/ICCV48922.2021.00986

Q. Sun, Y. Liu, L. Li, H. Sun, L. Zhao, and C. Tian, FA-Mamba: a Mamba-based approach for remote sensing change detection, Proceedings of the SPIE, 2025, Volume 13697.

10.1117/12.3076336

I. Corley, C. Robinson, and A. Ortiz, A change detection reality check. 2024. arXiv: 2402.06994.

Q. Shi, M. Liu, S. Li, X. Liu, F. Wang, and L. Zhang, A deeply supervised attention metric-based network and an open aerial image dataset for remote sensing change detection. IEEE Transactions on Geoscience and Remote Sensing. 60, 2021, pp. 1-16.

10.1109/TGRS.2021.3085870

L. Zhao, L. Wan, L. Ma, and Y. Zhang, HiSTENet: History-Integrated Spatial-Temporal Information Extraction Network for Time Series Remote Sensing Image Change Detection. Remote Sensing. 17(5), 2025, Article 792.

10.3390/rs17050792

Information

Publisher :Korean Academy of Space Security
Publisher(Ko) :한국우주안보학회
Journal Title :JOURNAL OF SPACE SECURITY
Journal Title(Ko) :한국우주안보학회지
Volume : 3
No :1
Pages :13-21
Received Date : 2026-03-04
Revised Date : 2026-05-29
Accepted Date : 2026-05-29
DOI :https://doi.org/10.23386/joss.2026.3.1.002

JOURNAL OF SPACE SECURITY ISSN:3058-5759(Print) 한국우주안보학회지

All Issue