Instance Segmentation of Buildings Using Keypoints

Qingyu Li, Lichao Mou, Yuansheng Hua, Yao Sun, Pu Jin, Yilei Shi, Xiao Xiang Zhu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

15 Scopus citations


Building segmentation is of great importance in the task of remote sensing imagery interpretation. However, the existing semantic segmentation and instance segmentation methods often lead to segmentation masks with blurred boundaries. In this paper, we propose a novel instance segmentation network for building segmentation in high-resolution remote sensing images. More specifically, we consider segmenting an individual building as detecting several keypoints. The detected keypoints are subsequently reformulated as a closed polygon, which is the semantic boundary of the building. By doing so, the sharp boundary of the building could be preserved. Experiments are conducted on selected Aerial Imagery for Roof Segmentation (AIRS) dataset, and our method achieves better performance in both quantitative and qualitative results with comparison to the state-of-the-art methods. Our network is a bottom-up instance segmentation method that could well preserve geometric details.

Original languageEnglish
Title of host publication2020 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2020 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Number of pages4
ISBN (Electronic)9781728163741
StatePublished - 26 Sep 2020
Externally publishedYes
Event2020 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2020 - Virtual, Waikoloa, United States
Duration: 26 Sep 20202 Oct 2020

Publication series

NameInternational Geoscience and Remote Sensing Symposium (IGARSS)


Conference2020 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2020
Country/TerritoryUnited States
CityVirtual, Waikoloa


  • aerial imagery
  • building
  • deep network
  • instance segmentation
  • keypoint detection


Dive into the research topics of 'Instance Segmentation of Buildings Using Keypoints'. Together they form a unique fingerprint.

Cite this