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

Global point cloud registration is essential in many robotics tasks like loopclosing and relocalization. Unfortunately, the registration often suffers fromthe low overlap between point clouds, a frequent occurrence in practicalapplications due to occlusion and viewpoint change. In this paper, we propose agraph-theoretic framework to address the problem of global point cloudregistration with low overlap. To this end, we construct a consistency graph tofacilitate robust data association and employ graduated non-convexity (GNC) forreliable pose estimation, following the state-of-the-art (SoTA) methods.  Unlike previous approaches, we use semantic cues to scale down the densepoint clouds, thus reducing the problem size. Moreover, we address theambiguity arising from the consistency threshold by constructing a pyramidgraph with multi-level consistency thresholds. Then we propose a cascadedgradient ascend method to solve the resulting densest clique problem and obtainmultiple pose candidates for every consistency threshold. Finally, fastgeometric verification is employed to select the optimal estimation frommultiple pose candidates. Our experiments, conducted on a self-collected indoordataset and the public KITTI dataset, demonstrate that our method achieves thehighest success rate despite the low overlap of point clouds and low semanticquality. We have open-sourced our codehttps://github.com/HKUST-Aerial-Robotics/Pagor for this project.

Pyramid Semantic Graph-based Global Point Cloud Registration with Low Overlap