Coupled structure-from-motion and 3D symmetry detection for urban facades

Abstract

EPFLRepeated structures are ubiquitous in urban facades. Such repetitions lead to ambiguity in establishing correspondences across sets of unordered images. A decoupled structure-from-motion reconstruction followed by symmetry detection often produces errors: outputs are either noisy and incomplete, or even worse, appear to be valid but actually have a wrong number of repeated elements. We present an optimization framework for extracting repeated elements in images of urban facades, while simultaneously calibrating the input images and recovering the 3D scene geometry using a graph-based global analysis. We evaluate the robustness of the proposed scheme on a range of challenging examples containing widespread repetitions and non-distinctive features. These image sets are common but cannot be handled well with state-of-the-art methods. We show that the recovered symmetry information along with the 3D geometry enables a range of novel image editing operations that maintain consistency across the images.

BigSUR

“…Furthermore, satellite data have lower resolution and drones can capture only smaller areas. While LiDAR produces point clouds directly, images must be processed to produce sparse [Snavely et al 2006] or dense [Ceylan et al 2013;Furukawa and Ponce 2010] point clouds. Some integrated modeling pipelines extract mass models from images directly [Dick et al 2004;Garcia-Dorado et al 2013;Vanegas et al 2010].…”

Mind Your Grey Tones – Examining the Influence of Decolourization Methods on Interest Point Extraction and Matching for Architectural Image-Based Modelling

“…As is usual with images from such cameras, they have the sRGB IEC61966-2.1 colour profile embedded. This image set is made publically available by Ceylan and colleagues (http://www.duygu-ceylan.com/duyguceylan/symmCalib.html) and accompanies their article which deals with the ambiguity in establishing correspondences encountered by SfM algorithms when dealing with repeated structures in urban facades (Ceylan et al, 2014). Whereas wellestablished SfM approaches often erroneously reconstruct such scenes, their framework allows for the robust extraction of those repeated façade elements and generates a proper output using a symmetry-based SfM algorithm.…”

Symmetrization of facade layouts

“…1 shows two such examples). In computer graphics, symmetry has become a key ingredient for facade modeling [2][3][4] , high level structural analysis [5] , and manipulation [6] . However, when digitizing existing symmetric facades (e.g., extraction of facade layouts), errors are unavoidably introduced during the digitization process.…”