Learning Dense Stereo Matching for Digital Surface Models from Satellite Imagery
This addresses a largely overlooked task in deep learning for satellite imagery, representing one of the first attempts in this domain, though it appears incremental as it builds on existing stereo reconstruction techniques.
The paper tackled the problem of generating Digital Surface Models from satellite imagery, which is difficult due to varying image pairs, by presenting a neural network and training scheme, resulting in smooth models that preserve boundaries and enable further processing.
Digital Surface Model generation from satellite imagery is a difficult task that has been largely overlooked by the deep learning community. Stereo reconstruction techniques developed for terrestrial systems including self driving cars do not translate well to satellite imagery where image pairs vary considerably. In this work we present neural network tailored for Digital Surface Model generation, a ground truthing and training scheme which maximizes available hardware, and we present a comparison to existing methods. The resulting models are smooth, preserve boundaries, and enable further processing. This represents one of the first attempts at leveraging deep learning in this domain.