Building Extraction at Scale using Convolutional Neural Network: Mapping of the United States

TL;DR

This study compares four CNN architectures for large-scale building footprint extraction from aerial imagery across the US, proposing enhancements like signed-distance labels and IR data integration to improve accuracy and instance-level mapping.

Contribution

It provides the first large-scale comparative analysis of CNNs for building extraction in the US and introduces a novel approach combining signed-distance labels with SegNet.

Findings

SegNet was identified as the best performing CNN architecture.

The proposed method achieved high precision and recall in large-scale building mapping.

Fusion of IR data improved building extraction accuracy.

Abstract

Establishing up-to-date large scale building maps is essential to understand urban dynamics, such as estimating population, urban planning and many other applications. Although many computer vision tasks has been successfully carried out with deep convolutional neural networks, there is a growing need to understand their large scale impact on building mapping with remote sensing imagery. Taking advantage of the scalability of CNNs and using only few areas with the abundance of building footprints, for the first time we conduct a comparative analysis of four state-of-the-art CNNs for extracting building footprints across the entire continental United States. The four CNN architectures namely: branch-out CNN, fully convolutional neural network (FCN), conditional random field as recurrent neural network (CRFasRNN), and SegNet, support semantic pixel-wise labeling and focus on capturing textural information at multi-scale. We use 1-meter resolution aerial images from National Agriculture Imagery Program (NAIP) as the test-bed, and compare the extraction results across the four methods. In addition, we propose to combine signed-distance labels with SegNet, the preferred CNN architecture identified by our extensive evaluations, to advance building extraction results to instance level. We further demonstrate the usefulness of fusing additional near IR information into the building extraction framework. Large scale experimental evaluations are conducted and reported using metrics that include: precision, recall rate, intersection over union, and the number of buildings extracted. With the improved CNN model and no requirement of further post-processing, we have generated building maps for the United States. The quality of extracted buildings and processing time demonstrated the proposed CNN-based framework fits the need of building extraction at scale.