A Cross-Domain Survey of Medical and Remote Sensing Image Registration Methods
Year: 2026
Type: article
Abstract: Image registration is central to multimodal fusion, longitudinal analysis, and spatial correspondence in both medical imaging and remote sensing, yet these two literatures have largely evolved in parallel despite sharing core algorithmic challenges. This survey reviews more than 200 representative studies across medical, remote-sensing, and general or cross-domain registration settings to provide a unified view of how registration methods are designed, evaluated, and deployed across domains. To our knowledge, this is the first survey to organize medical and remote-sensing registration within a single cross-domain taxonomy spanning acquisition settings, transformation models, and algorithmic families while also comparing datasets, evaluation practice, metric limitations, failure modes, and reproducibility considerations. We show that the two domains often prioritize different strengths: medical registration emphasizes anatomical consistency and topology-preserving deformation, whereas remote sensing emphasizes robustness to sensing heterogeneity, scale variation, and large-scene misalignment. By identifying these complementary strengths, the survey highlights concrete opportunities for cross-domain transfer, including stronger geometric priors for remote sensing and more scalable multimodal feature learning for medical imaging. The survey concludes with open challenges in benchmarking, robustness, and foundation-model integration for future registration systems.
Source: IEEE Access
Institution: Duke University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,075
OpenAlex ID: W7212193759