Reducing patch-like Errors in SAR offset tracking displacements using logarithmic transformation and a weighted NCC algorithm

ABSTRACT

Pixel offset tracking (OT) algorithm is a useful tool for measuring large surface displacements by matching amplitudes in master and slave synthetic aperture radar (SAR) images. However, strong backscatters can cause homogeneous errors within a matching window (referred to as patch-like errors) in traditional OT processing, thereby misleading the interpretation of displacement events, especially over a small area. In this letter, we proposed an improved SAR OT algorithm to reduce patch-like errors. In which, a logarithmic transformation was firstly utilized to narrow the SAR amplitude range between strong and weak back scatterers. Strong backscatters causing patch-like errors were then statistically detected with an indicator of median absolute deviation. Finally, those strong backscatters were excluded from SAR OT processing using a weighted normalized cross-correlation scheme, in order to reduce the caused patch-like errors. Two real data tests over the Shuozhou and Yulin coal mining areas, China, suggest that the mean accuracy of the displacements estimated by the presented method improved about 30%, with respect to that estimated by the traditional OT algorithm. The proposed SAR OT algorithm offers a robust option to measure large displacements, especially over a small area, associated with anthropologic or geophysical activities.

KEYWORDS:

• large displacement
• patch-like errors
• SAR offset tracking
• logarithmic transformation
• coal mining subsidence

Acknowledgments

The authors would like to thank the anonymous reviewers for the insightful suggestions and the German Aerospace Center (DLR) for providing the TerraSAR-X images of the study area (No. MTH3273).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

We also thank the anonymous reviewers for the constructive suggestions. This work was supported in part by the National Natural Science Foundation of China under Grant Nos. 42274057 and 41904005, the Science and Technology Innovation Program of Hunan Province under Grant No. 2021RC3008, the Research Foundation of Education Bureau of Hunan Province, China under Grant No. 20K134, Central South University Innovation-Driven Research Programme under Grant No. 2023CXQD006, and Key Laboratory of Natural Resources Monitoring and Supervision in Southern Hilly Region, Ministry of Natural Resources under Grant No. NRMSSHR2022Y09.