Map-Repair: Deep Cadastre Maps Alignment and Temporal Inconsistencies Fix in Satellite Images

TL;DR

This paper introduces a deep learning method for aligning and correcting inconsistencies in cadastre maps using satellite images, effectively handling misalignments and label noise to improve map accuracy.

Contribution

The work presents an end-to-end deep learning approach that corrects both label noise and spatial misalignments in cadastre maps from satellite imagery, enhancing map updating processes.

Findings

Robustness to severe misalignments demonstrated

Effective correction of label noise in cadastre maps

Potential application in OpenStreetMap updates

Abstract

In the fast developing countries it is hard to trace new buildings construction or old structures destruction and, as a result, to keep the up-to-date cadastre maps. Moreover, due to the complexity of urban regions or inconsistency of data used for cadastre maps extraction, the errors in form of misalignment is a common problem. In this work, we propose an end-to-end deep learning approach which is able to solve inconsistencies between the input intensity image and the available building footprints by correcting label noises and, at the same time, misalignments if needed. The obtained results demonstrate the robustness of the proposed method to even severely misaligned examples that makes it potentially suitable for real applications, like OpenStreetMap correction.