Rotated Rectangles for Symbolized Building Footprint Extraction

CVPRW 2018 pp. 225-228

doi:10.1109/CVPRW.2018.00039 /cvprw/2018/dickenson2018cvprw-rotated/

Abstract

Building footprints (BFP) provide useful visual context for users of digital maps when navigating in space. This paper proposes a method for extracting and symbolizing building footprints from satellite imagery using a convolutional neural network (CNN). The CNN architecture outputs rotated rectangles, providing a symbolized approximation that works well for small buildings. Experiments are conducted on the four cities in the DeepGlobe Challenge dataset (Las Vegas, Paris, Shanghai, Khartoum). Our method performs best on suburbs consisting of individual houses. These experiments show that either large buildings or buildings without clear delineation produce weaker results in terms of precision and recall.

CVPRW Semantic Scholar

Cite

Text

Dickenson and Gueguen. "Rotated Rectangles for Symbolized Building Footprint Extraction." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2018. doi:10.1109/CVPRW.2018.00039

Markdown

[Dickenson and Gueguen. "Rotated Rectangles for Symbolized Building Footprint Extraction." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2018.](https://mlanthology.org/cvprw/2018/dickenson2018cvprw-rotated/) doi:10.1109/CVPRW.2018.00039

BibTeX

@inproceedings{dickenson2018cvprw-rotated,
  title     = {{Rotated Rectangles for Symbolized Building Footprint Extraction}},
  author    = {Dickenson, Matt and Gueguen, Lionel},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2018},
  pages     = {225-228},
  doi       = {10.1109/CVPRW.2018.00039},
  url       = {https://mlanthology.org/cvprw/2018/dickenson2018cvprw-rotated/}
}