English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Structure-from-Motion (SfM) aims to recover 3D scene structures and cameraposes based on the correspondences between input images, and thus the ambiguitycaused by duplicate structures (i.e., different structures with strong visualresemblance) always results in incorrect camera poses and 3D structures. Todeal with the ambiguity, most existing studies resort to additional constraintinformation or implicit inference by analyzing two-view geometries or featurepoints. In this paper, we propose to exploit high-level information in thescene, i.e., the spatial contextual information of local regions, to guide thereconstruction. Specifically, a novel structure is proposed, namely,{\textit{track-community}}, in which each community consists of a group oftracks and represents a local segment in the scene. A community detectionalgorithm is used to partition the scene into several segments. Then, thepotential ambiguous segments are detected by analyzing the neighborhood oftracks and corrected by checking the pose consistency. Finally, we performpartial reconstruction on each segment and align them with a novelbidirectional consistency cost function which considers both 3D-3Dcorrespondences and pairwise relative camera poses. Experimental resultsdemonstrate that our approach can robustly alleviate reconstruction failureresulting from visually indistinguishable structures and accurately merge thepartial reconstructions.

TC-SfM: Robust Track-Community-Based Structure-from-Motion