/building-footprints-update

An update algorithm without manual relabeling is proposed to address the problem regarding differences between the data distributions of pre- and post-temporal images in the building update process

Primary LanguagePython

Building footprints Update

This code is implemented for the article "Learning color distributions from bitemporal remote sensing images to update existing building footprints"

image-color-translation

The difference between the data distributions of different time-phase images is reduced using the image color translation method

semantic-segmentation

Semantic segmentation model predicts the segmentation results of the images from the latest period

post-processing-update-strategy

Strictly retain the existing labels of unchanged regions to attain the updated results