Skip to Main content Skip to Navigation
Conference papers

Polygonal Building Segmentation by Frame Field Learning

Abstract : While state of the art image segmentation models typically output segmentations in raster format, applications in geographic information systems often require vector polygons. To help bridge the gap between deep network output and the format used in downstream tasks, we add a frame field output to a deep segmentation model for extracting buildings from remote sensing images. We train a deep neural network that aligns a predicted frame field to ground truth contours. This additional objective improves segmentation quality by leveraging multi-task learning and provides structural information that later facilitates polygonization; we also introduce a polygonization algorithm that utilizes the frame field along with the raster segmentation. Our code is available at https://github.com/Lydorn/Polygonization-by-Frame-Field-Learning.
Complete list of metadata

https://hal.inria.fr/hal-02548545
Contributor : Nicolas Girard <>
Submitted on : Wednesday, March 31, 2021 - 2:05:40 PM
Last modification on : Friday, April 2, 2021 - 3:31:07 AM

File

archive.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02548545, version 2

Citation

Nicolas Girard, Dmitriy Smirnov, Justin Solomon, Yuliya Tarabalka. Polygonal Building Segmentation by Frame Field Learning. CVPR 2021 - IEEE Conference on Computer Vision and Pattern Recognition, Jun 2021, Pittsburg / Virtual, United States. ⟨hal-02548545v2⟩

Share

Metrics

Record views

55

Files downloads

301