Towards Large-Scale City Reconstruction from Satellites
Automatic city modeling from satellite imagery is one of the biggest challenges in urban reconstruction. Existing methods produce at best rough and dense Digital Surface Models. Inspired by recent works on semantic 3D reconstruction and region-based stere
- PDF / 10,236,457 Bytes
- 16 Pages / 439.37 x 666.142 pts Page_size
- 43 Downloads / 168 Views
Abstract. Automatic city modeling from satellite imagery is one of the biggest challenges in urban reconstruction. Existing methods produce at best rough and dense Digital Surface Models. Inspired by recent works on semantic 3D reconstruction and region-based stereovision, we propose a method for producing compact, semantic-aware and geometrically accurate 3D city models from stereo pair of satellite images. Our approach relies on two key ingredients. First, geometry and semantics are retrieved simultaneously bringing robustness to occlusions and to low image quality. Second, we operate at the scale of geometric atomic region which allows the shape of urban objects to be well preserved, and a gain in scalability and efficiency. We demonstrate the potential of our algorithm by reconstructing different cities around the world in a few minutes. Keywords: 3D reconstruction Urban scenes
1
·
City modeling
·
Satellite imagery
·
Introduction
Automatic city modeling has received an increasing interest during the last decade. In applicative fields such as urban planning, telecommunications and disaster control, producing compact and accurate 3D models is crucial. Aerial acquisitions with Lidar scanning or multi-view imagery constitute the best way so far to automatically create 3D models on large-scale urban scenes [1]. Because of high acquisition costs and authorization constraints, aerial acquisitions are, however, restricted to some spotlighted cities in the world. In particular, Geographic Information System (GIS) companies propose catalogs with typically a few hundred cities in the world. Satellite imagery exhibits higher potential with lower costs, a worldwide coverage and a high acquisition frequency. Satellites have however several technical restrictions that prevent GIS practitioners from producing compact city models in an automatic way [2]. Inspired by recent works on semantic 3D reconstruction and region-based stereovision, we propose a method for producing compact, semantic-aware and Electronic supplementary material The online version of this chapter (doi:10. 1007/978-3-319-46454-1 6) contains supplementary material, which is available to authorized users. c Springer International Publishing AG 2016 B. Leibe et al. (Eds.): ECCV 2016, Part V, LNCS 9909, pp. 89–104, 2016. DOI: 10.1007/978-3-319-46454-1 6
90
L. Duan and F. Lafarge
Fig. 1. Reconstruction of Denver downtown. Starting from a stereo pair of satellite images (left), our algorithm produces a compact and semantic-aware 3D model (right) in a few minutes
geometrically accurate 3D city models from stereo pairs of satellite images. Our approach relies on two key ingredients. First, geometry and semantics are retrieved simultaneously bringing robustness to occlusions and to low image quality. Second, contrary to pixel-based methods, we operate at the scale of geometric atomic region: it allows the shape of urban objects to be better preserved, and also a gain in scalability and efficiency. Figure 1 illustrates our goal.
2
Related Works
Our review
Data Loading...
