Lidar SLAM Algorithm Based on Online Point Cloud Removal in Pseudo Occupied Area

Fig. 5. Sequence 00 and sequence 07 dynamic point cloud removal effect for different algorithms. (a) Original iamges; (b) Removert algorithm; (c) ERASOR algorithm; (d) DOR-LOAM algorithm

Download full size

Fig. 6. Detail of point cloud removal effect of sequence 05 for different algorithms. (a) Ground truth; (b) ERASOR algorithm; (c) DOR-LOAM algorithm

Download full size

Fig. 7. Comparison of tracks from sequence 00 to sequence 09

Download full size

Fig. 8. Comparison of dynamic removal rate parameters

Download full size

Fig. 9. Comparison of running time of each module

Download full size

Table 1. Comparison of dynamic point cloud removal accuracy for different algorithms

View table

Table 1. Comparison of dynamic point cloud removal accuracy for different algorithms

Sequence	Evaluation index	Removert	ERASOR	DOR-LOAM
00	$R_{P}$	85.50	93.98	97.29
	$R_{R}$	99.35	97.08	97.06
	$F_{1}$	91.90	95.50	97.18
01	$R_{P}$	94.22	91.49	96.45
	$R_{R}$	93.61	95.38	94.87
	$F_{1}$	93.91	93.40	95.65
02	$R_{P}$	76.32	87.73	84.27
	$R_{R}$	96.80	97.01	97.30
	$F_{1}$	85.35	92.10	90.32
05	$R_{P}$	86.90	88.73	95.33
	$R_{R}$	87.88	98.26	97.53
	$F_{1}$	87.39	93.30	96.42
07	$R_{P}$	80.69	90.62	97.30
	$R_{R}$	98.82	99.27	98.80
	$F_{1}$	88.84	94.80	98.04
Average	$R_{P}$	84.726	90.51	94.13
	$R_{R}$	95.292	97.40	97.11
	$F_{1}$	89.478	93.82	95.52

Table 2. Accuracy comparison of different algorithms

View table

Table 2. Accuracy comparison of different algorithms

Sequence	E-LOAM		T-LOAM		NDT-LOAM	PSF-LO	SuMa++		DOR-LOAM
	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{r o t}$ / ［（°）·m^-1］	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{r o t}$ / ［（°）·m^-1］	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{r o t}$ / ［（°）·m^-1］	${\bar{E}}_{t r a n s}$ /%	${\bar{E}}_{r o t}$ / ［（°）·m^-1］
00	1.17	0.0047	0.98	0.0060	0.79	0.64	0.64	0.0022	0.82	0.0028
01	2.92	0.0062	2.09	0.0052	1.46	1.32	1.60	0.0046	1.83	0.0050
02	2.36	0.0082	1.01	0.0039	1.09	0.87	1.00	0.0037	1.11	0.0036
03	1.16	0.0068	1.10	0.0082	0.65	0.75	0.67	0.0046	0.65	0.0043
04	1.39	0.0050	0.68	0.0068	0.31	0.66	0.37	0.0026	0.59	0.0038
05	0.82	0.0036	0.55	0.0032	0.54	0.45	0.40	0.0020	0.54	0.0023
06	1.34	0.0057	0.56	0.0031	0.56	0.47	0.46	0.0021	0.48	0.0023
07	1.21	0.0066	0.50	0.0047	0.27	0.46	0.34	0.0019	0.37	0.0025
08	1.63	0.0063	0.94	0.0033	1.04	0.94	1.10	0.0035	0.93	0.0030
09	1.36	0.0057	0.80	0.0040	0.74	0.56	0.47	0.0023	0.71	0.0028
10	1.84	0.0063	1.12	0.0061	1.12	0.54	0.66	0.0028	0.92	0.0037
Average	1.56	0.0059	0.93	0.0049	0.90	0.74	0.70	0.0029	0.81	0.0033

Table 3. Comparison of ablation experiments and different algorithms
View table
Table 3. Comparison of ablation experiments and different algorithms
Algorithm Baseline Baseline+DOR E-LOAM T-LOAM NDT-LOAM
${\bar{E}}_{t r a n s}$ /% 1.28 0.81 1.56 0.93 0.90
${\bar{E}}_{r o t}$ /
［（°）·m^-1］
0.0044 0.0033 0.0059 0.0049 -
Average time /ms 76.33 87.48 77.38 100 -

Tools

Get Citation

Copy Citation Text

Qingxuan Zeng, Qiang Li, Weizhi Nie. Lidar SLAM Algorithm Based on Online Point Cloud Removal in Pseudo Occupied Area[J]. Acta Optica Sinica, 2023, 43(20): 2028001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites