Resnet-Unet considering Patches (RUP) network to solve the problem of patches due to shadows in extracting building top information

ABSTRACT

Some highrise buildings have elevator rooms or stair houses on the top, which will lead to obvious shadows on the top of the building. When using the deeplearning algorithm to extract the top information of buildings on highresolution remote sensing images, the extracted results have breakage or holes, which leads to the appearance of patches. Based on the Unet and Resnet network, the ResnetUnet considering Patches (RUP) Network is proposed. The improved model not only has high accuracy but also solves the problems of incomplete extraction and patches when there are shadows on the top of the building.

KEYWORDS:

• Highresolution remote sensing image
• building information extraction
• shadows on the top of buildings
• patches
• deep learning

Acknowledgments

The authors wish to thank the editors and reviewers.

Disclosure statement

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

Additional information

Funding

The National Science and Technology Basic Resources Survey Project [2019FY202504].