• Acta Optica Sinica
  • Vol. 43, Issue 24, 2428003 (2023)
Yongjian Fu, Zongchun Li*, Hua He, Li Wang, and Cong Li
Author Affiliations
  • Institute of Geospatial Information, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, Henan , China
  • show less
    DOI: 10.3788/AOS230548 Cite this Article Set citation alerts
    Yongjian Fu, Zongchun Li, Hua He, Li Wang, Cong Li. HRegNet-LO: LiDAR Odometry Measurement Based on End-to-End Deep Neural Network[J]. Acta Optica Sinica, 2023, 43(24): 2428003 Copy Citation Text show less
    Flow chart of HRegNet-LO algorithm
    Fig. 1. Flow chart of HRegNet-LO algorithm
    Feature points extraction result of a point cloud
    Fig. 2. Feature points extraction result of a point cloud
    Flow chart of HRegNet registration network[19]
    Fig. 3. Flow chart of HRegNet registration network[19]
    Illustration of feature points matching. (a) Point set T1nFn; (b) point set Fmap
    Fig. 4. Illustration of feature points matching. (a) Point set T1nFn; (b) point set Fmap
    Visualization of results of proposed algorithm on testing datasets along each coordinate axis. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fig. 5. Visualization of results of proposed algorithm on testing datasets along each coordinate axis. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Aerial view of results of proposed algorithm on testing datasets. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fig. 6. Aerial view of results of proposed algorithm on testing datasets. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fixed distance relative errors of three algorithms on testing datasets. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fig. 7. Fixed distance relative errors of three algorithms on testing datasets. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fixed distance relative errors of algorithms in ablation experiments. (a) Seq 08; (b) Seq 09; (c) Seq 10
    Fig. 8. Fixed distance relative errors of algorithms in ablation experiments. (a) Seq 08; (b) Seq 09; (c) Seq 10
    SeqFrameLength /mMax speed /(km·h-1SeqFrameLength /mMax speed /(km·h-1
    004541372446061101123251
    01110124539607110169439
    024661506749084071322243
    0380156031091591170552
    042713935610120191951
    052761220540
    Table 1. Details of experimental datasets
    AlgorithmAccuracy and time consumingSeq 08Seq 09Seq 10RMSE
    LOAMRRE /[(°)·m-10.00410.00310.00320.0035
    RTE /%1.05690.73261.07060.9661
    Time consuming /ms878385
    F-LOAMRRE /[(°)·m-10.00480.00510.00600.0053
    RTE /%1.35321.25201.52181.3802
    Time consuming /ms767475
    HRegNet-LORRE /[(°)·m-10.00330.00310.00390.0035
    RTE /%1.04720.71901.04800.9508
    Time consuming /ms959796
    Table 2. Accuracy and time consuming of three algorithms on testing datasets
    AlgorithmAccuracySeq 08Seq 09Seq 10RMSE
    ICP-odomRRE /[(°).m-10.06210.11770.05860.0840
    RTE /%14.513631.025116.745022.0121
    HRegNet-odomRRE /[(°).m-10.04020.01610.02120.0278
    RTE /%9.83443.72634.55246.6163
    LOAM-odomRRE /[(°).m-10.02110.01910.01820.0195
    RTE /%4.97155.98073.82405.0036
    HRegNet-LO-odomRRE /[(°).m-10.01160.00990.01120.0109
    RTE /%4.14012.44752.45553.1177
    Table 3. Results of ablation experiments
    Yongjian Fu, Zongchun Li, Hua He, Li Wang, Cong Li. HRegNet-LO: LiDAR Odometry Measurement Based on End-to-End Deep Neural Network[J]. Acta Optica Sinica, 2023, 43(24): 2428003
    Download Citation