A hierarchical spatial-temporal graph-kernel for high-resolution SAR image change detection

ABSTRACT

Effective utilization of structural information is important for high-resolution synthetic aperture radar (SAR) image change detection. For comprehensively utilizing the local and global structures in SAR images, a hierarchical spatial-temporal graph kernel (STGK) method is proposed in this paper for high-resolution SAR image change detection. First, the bi-temporal hierarchical graph models are constructed for extracting the local-global structures in the bi-temporal SAR images. Then, a STGK function, which measures the spatial and temporal similarities between the local-global structures, is constructed for indicating the change levels between the bi-temporal images. Finally, a support vector machine (SVM) is implemented with the STGK function for producing the final change detection results. Experimental results on real GaoFen-3 SAR data sets demonstrate the effectiveness of the proposed method, and prove that the STGK method is capable of detecting changed areas with relatively complex structures.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (61701154 and 61701157), the Natural Science Foundation of Anhui Province (1808085QF185), the China Postdoctoral Science Foundation (2018M630703). Moreover, the authors would like to thank the reviewers and the editor for the constructive comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the China Postdoctoral Science Foundation [2018M630703]; National Natural Science Foundation of China [61701154,61701157]; Natural Science Foundation of Anhui Province [1808085QF185].