Lidar scan matching with rtk-gnss positioning and geometric constraints

Masafumi Nakagawa, Shinjiro Abe, Sho Sanuka, Kazuo Saito, Masahiro Miyo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Scan matching in simultaneous localization and mapping (SLAM) has several technical issues, such as error accumulation, high processing cost in point cloud matching and optimization, and position loss problems after scan matching failures. Error adjustment processing can improve the performance of SLAM with loop closure and global optimization approach. However, measurement path plans and higher processing cost are required for the error adjustment and global optimization. In contrast, global navigation satellite system (GNSS) positioning can simplify the scan matching. Thus, we propose scan matching for multilayer LiDAR data registration with RTK-GNSS positioning and geometric constraints. Through experiments on point cloud acquisition with multilayer LiDAR and a single-frequency RTK-GNSS positioning device, we verify that our methodology can integrate point clouds acquired in mobile mapping without an inertial measurement unit. We also confirm that our methodology can avoid error accumulation problems in conventional SLAM processing.

Original languageEnglish
Title of host publicationACRS 2020 - 41st Asian Conference on Remote Sensing
PublisherAsian Association on Remote Sensing
ISBN (Electronic)9781713829089
Publication statusPublished - 2020
Event41st Asian Conference on Remote Sensing, ACRS 2020 - Deqing City, Virtual, China
Duration: 2020 Nov 92020 Nov 11

Publication series

NameACRS 2020 - 41st Asian Conference on Remote Sensing

Conference

Conference41st Asian Conference on Remote Sensing, ACRS 2020
Country/TerritoryChina
CityDeqing City, Virtual
Period20/11/920/11/11

Keywords

  • Geometric constraints
  • Mobile mapping
  • Multilayer LiDAR
  • RTK-GNSS
  • Scan matching

ASJC Scopus subject areas

  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Lidar scan matching with rtk-gnss positioning and geometric constraints'. Together they form a unique fingerprint.

Cite this