Author Affiliations
1Hubei Provincial Research Institute of Land and Resources, Wuhan, Hubei 430071, China2Hunan Glonavin Information Technology Co., Ltd., Changsha, Hunan 410006, Chinashow less
Fig. 1. Algorithm framework diagram
Fig. 2. Pose graph integrated with GNSS position data
Fig. 3. Accuracy of point cloud registration pose estimation
Fig. 4. GNSS-SLAM pose correction when there is no loopback
Fig. 5. Diagram of loop closure optimization
Fig. 6. Loop closure detection diagram
Fig. 7. Hardware picture of experiment
Fig. 8. Test without loopback in urban environment. (a) With GNSS pose optimization; (b) without GNSS pose optimization
Fig. 9. Loop trajectory test in urban environment. (a) With GNSS pose optimization; (b) without GNSS pose optimization
Fig. 10. Test without loopback in non-urban environment. (a) With GNSS pose optimization; (b) without GNSS pose optimization
Fig. 11. Picture of non-urban environment
Fig. 12. Loop trajectory test in non-urban environment. (a) With GNSS pose optimization; (b) without GNSS pose optimization
Hardware | Software |
---|
Sensor | Models (parameters) | Name | Version | LiDAR | Robosense RS-LiDAR-16(ranging accuracy ±0.02 m,distance measuring range of 100-150 m) | Operating System | Ubuntu16.04 LTS | GPS module | Glonavin GNSS Single site location module(Horizontal position precision 1 m,vertical position precision 2 m) | ROS(robot operating system) | Kinetic | Hardware platform | CPU (Intel i7 8core 1.8 GHZ),RAM(16 GB) | Point cloud registrationalgorithm | ICP |
|
Table 1. Experimental software and hardware related instructions and parameters
Experiment | With GNSS poseoptimization | Without GNSSpose optimization |
---|
δd | δy | δz | δd | δy | δz |
---|
Withoutloopback | Urban | 1.11 | 0.07 | 0.97 | 3.83 | 0.35 | 3.56 | Non-urban | 0.62 | 0.17 | 0.09 | 25.25 | 14.30 | 3.32 | Loop trajectory | Urban | | First loop drift | 0.20 | 0.12 | 0.15 | 30.21 | 8.98 | 15.38 | Second loop drift | 0.11 | 0.10 | 0.08 | 51.95 | 10.30 | 39.76 | Non-urban | | First loop drift | 0.16 | 0.14 | 0.05 | 5.52 | 4.92 | 0.33 | Second loop drift | 0.10 | 0.09 | 0.04 | 77.49 | 19.30 | 55.75 |
|
Table 2. Trajectory drift comparisonm