• Laser & Optoelectronics Progress
  • Vol. 58, Issue 24, 2428003 (2021)
Dong Shen, Yuhang Xu*, Qiang Li, Jing Di, and Xia Huang
Author Affiliations
  • School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou, Gansu 730070, China
  • show less
    DOI: 10.3788/LOP202158.2428003 Cite this Article Set citation alerts
    Dong Shen, Yuhang Xu, Qiang Li, Jing Di, Xia Huang. Research on Laser SLAM Algorithm Based on Sparse Pose Optimization[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2428003 Copy Citation Text show less
    Graph-based SLAM framework
    Fig. 1. Graph-based SLAM framework
    Compressed column storage format
    Fig. 2. Compressed column storage format
    Flow chart of sparse matrix establishment
    Fig. 3. Flow chart of sparse matrix establishment
    Continuable LM algorithm
    Fig. 4. Continuable LM algorithm
    Overall hardware structure diagram of the system
    Fig. 5. Overall hardware structure diagram of the system
    The main parts and physical maps of the mobile robot
    Fig. 6. The main parts and physical maps of the mobile robot
    Experimental results of the three algorithms for constructing map. (a) Gmapping; (b) Hector; (c) proposed algorithm
    Fig. 7. Experimental results of the three algorithms for constructing map. (a) Gmapping; (b) Hector; (c) proposed algorithm
    Error curves of the three algorithms in corridor
    Fig. 8. Error curves of the three algorithms in corridor
    Error curves of the three algorithms in complex laboratory
    Fig. 9. Error curves of the three algorithms in complex laboratory
    ParameterGmappingHectorProposed algorithm
    Max error /m0.07900.093000.0770
    Min error /m0.02500.032000.0280
    Mean square error(MSE) /m20.00230.018700.0023
    Root mean square error(RMSE) /m0.04820.136970.0481
    Mean absolute error(MAE) /m0.04430.051400.0452
    Table 1. Error analysis of the three algorithms in the corridor environment
    ParameterGmappingHectorProposed algorithm
    Max error /m0.03100.05200.0250
    Min error /m0.01700.02400.0130
    MSE /m20.00050.00110.0003
    RMSE /m0.02310.03430.0194
    MAE /m0.02270.03290.0190
    Table 2. Error analysis of the three algorithms in complex laboratory environment
    ParameterGmappingHectorProposed algorithm
    Mean CPU load /%7.07386.12135.3897
    Median CPU load /%5.61005.93105.3010
    Standard deviation of CPU load /%4.43121.9871.310
    Table 3. CPU load of the three algorithms
    EnvironmentGmappingHectorProposed algorithm
    Corridor25.8128.6424.40
    Laboratory105.92120.4898.04
    Table 4. Running time of the three algorithms unit: s
    Dong Shen, Yuhang Xu, Qiang Li, Jing Di, Xia Huang. Research on Laser SLAM Algorithm Based on Sparse Pose Optimization[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2428003
    Download Citation