Deep cross-domain building extraction for selective depth estimation from oblique aerial imagery

TL;DR

This paper presents a method combining transfer learning with CNNs for real-time building extraction and depth estimation from oblique aerial imagery, enhancing urban analysis capabilities.

Contribution

It introduces a novel approach integrating Faster R-CNN with semi-global matching for selective, real-time building reconstruction from aerial images.

Findings

Achieved 80% average precision in building extraction.

Demonstrated effective transfer learning from ground-based to aerial datasets.

Enabled real-time multi-view depth estimation with preserved object boundaries.

Abstract

With the technological advancements of aerial imagery and accurate 3d reconstruction of urban environments, more and more attention has been paid to the automated analyses of urban areas. In our work, we examine two important aspects that allow live analysis of building structures in city models given oblique aerial imagery, namely automatic building extraction with convolutional neural networks (CNNs) and selective real-time depth estimation from aerial imagery. We use transfer learning to train the Faster R-CNN method for real-time deep object detection, by combining a large ground-based dataset for urban scene understanding with a smaller number of images from an aerial dataset. We achieve an average precision (AP) of about 80% for the task of building extraction on a selected evaluation dataset. Our evaluation focuses on both dataset-specific learning and transfer learning. Furthermore, we present an algorithm that allows for multi-view depth estimation from aerial imagery in real-time. We adopt the semi-global matching (SGM) optimization strategy to preserve sharp edges at object boundaries. In combination with the Faster R-CNN, it allows a selective reconstruction of buildings, identified with regions of interest (RoIs), from oblique aerial imagery.