Paper Info
Title
An Integrated GNSS/INS/LiDAR-SLAM Positioning Method for Highly Accurate Forest Stem Mapping.
Abstract
Forest mapping, one of the main components of performing a forest inventory, is an important driving force in the development of laser scanning. Mobile laser scanning (MLS), in which laser scanners are installed on moving platforms, has been studied as a convenient measurement method for forest mapping in the past several years. Positioning and attitude accuracies are important for forest mapping using MLS systems. Inertial Navigation Systems (INSs) and Global Navigation Satellite Systems (GNSSs) are typical and popular positioning and attitude sensors used in MLS systems. In forest environments, because of the loss of signal due to occlusion and severe multipath effects, the positioning accuracy of GNSS is severely degraded, and even that of GNSS/INS decreases considerably. Light Detection and Ranging (LiDAR)-based Simultaneous Localization and Mapping (SLAM) can achieve higher positioning accuracy in environments containing many features and is commonly implemented in GNSS-denied indoor environments. Forests are different from an indoor environment in that the GNSS signal is available to some extent in a forest. Although the positioning accuracy of GNSS/INS is reduced, estimates of heading angle and velocity can maintain high accurate even with fewer satellites. GNSS/INS and the LiDAR-based SLAM technique can be effectively integrated to form a sustainable, highly accurate positioning and mapping solution for use in forests without additional hardware costs. In this study, information such as heading angles and velocities extracted from a GNSS/INS is utilized to improve the positioning accuracy of the SLAM solution, and two information-aided SLAM methods are proposed. First, a heading angle-aided SLAM (H-aided SLAM) method is proposed that supplies the heading angle from GNSS/INS to SLAM. Field test results show that the horizontal positioning accuracy of an entire trajectory of 800 m is 0.13 m and is significantly improved (by 70%) compared to that of a traditional GNSS/INS; second, a more complex information added SLAM solution that utilizes both heading angle and velocity information simultaneously (HV-aided SLAM) is investigated. Experimental results show that the horizontal positioning accuracy can reach a level of six centimetres with the HV-aided SLAM, which is a significant improvement (by 86%). Thus, a more accurate forest map is obtained by the proposed integrated method.
Year
DOI
Venue
2017
10.3390/rs9010003
REMOTE SENSING
Keywords
Field
DocType
forest mapping,MLS,GNSS/INS,LIDAR,SLAM,integration
Multipath propagation,Inertial navigation system,Computer vision,Laser scanning,Remote sensing,Forest inventory,Ranging,Lidar,Artificial intelligence,GNSS applications,Simultaneous localization and mapping,Geology
Journal
Volume
Issue
Citations 
9
1
6
PageRank 
References 
Authors
0.48
14
10
Name
Order
Citations
PageRank
Chuang Qian1112.63
Hui Liu261.49
Jian Tang3132259.93
Yuwei Chen412923.23
Harri Kaartinen560863.10
Antero Kukko648347.44
lingli zhu71069.87
Xinlian Liang819323.72
Liang Chen924742.33
Juha Hyyppä1043966.75