Low-complexity adaptive radius outlier removal filter based on PCA for lidar point cloud denoising

Yao Duan, Chuanchuan Yang, and Hongbin Li

Not Accessible

Your library or personal account may give you access

Get PDF
Email
Share
Get Citation
Copy Citation Text
Yao Duan, Chuanchuan Yang, and Hongbin Li, "Low-complexity adaptive radius outlier removal filter based on PCA for lidar point cloud denoising," Appl. Opt. 60, E1-E7 (2021)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Citation alert
Save article

Abstract

In autonomous driving, cars rely on light detection and ranging (lidar) to navigate the surroundings, but interference from the environment makes it difficult to retrieve useful information. To address this problem, this paper develops a noise reduction method to filter lidar point clouds (i.e., an adaptive radius outlier removal filter based on principal component analysis). We believe this method can outperform existing clustering algorithms when applied to point cloud images captured at a large distance from the lidar. Compared to traditional methods, the proposed method has higher precision and recall with an F-score up to 0.876 and complexity reduced by at least 50%.

Full Article | PDF Article

More Like This

TDNet: transformer-based network for point cloud denoising

Xueli Xu, Guohua Geng, Xin Cao, Kang Li, and Mingquan Zhou
Appl. Opt. 61(6) C80-C88 (2022)

Multi-level height maps-based registration method for sparse LiDAR point clouds in an urban scene

Bin Fang, Jie Ma, Pei An, Zhao Wang, Jun Zhang, and Kun Yu
Appl. Opt. 60(14) 4154-4164 (2021)

PDC-Net: Robust point cloud registration using deep cyclic neural network combined with PCA

Dengzhi Liu, Yu Zhang, Lin Luo, Jinlong Li, and Xiaorong Gao
Appl. Opt. 60(11) 2990-2997 (2021)

Previous Article Next Article

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Figures (4)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Tables (6)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Equations (12)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Filter	Complexity ( $+$ )	Complexity ( $\times$ )
SOR filter	$O (m^{2} l o g m)$	$O (m^{2} l o g m)$
ROR filter	$O (9 m^{2})$	$O (3 m^{2})$
Density-based clustering method	$O (13 m^{2})$	$O (1.5 m^{2})$
PCAAC method	$O (10 m^{2})$	$O (m^{2})$
PCAAR method	$O (6 m^{2})$	$O (2 m^{2})$

Filter	Time (Data Set 1)	Time (Data Set 2)
SOR filter	106 s	23.5 s
ROR filter	64.5 s	14.2 s
Density-based clustering method	24.6 s	6.75 s
PCAAC method	8.04 s	1.58 s
PCAAR method	16.7 s	4.98 s

Filter	Accuracy	Error	Precision	Recall	$F_{1}$
SOR filter	99.45%	0.55%	20.90%	32.98%	0.256
Two-stage SOR filter	99.23%	0.77%	21.02%	63.06%	0.315
ROR filter	99.61%	0.39%	43.66%	100%	0.608
AROR filter	99.76%	0.24%	55.32%	90.50%	0.686
Density-based clustering method	99.79%	0.21%	62.65%	76.18%	0.687
PCAAC method	99.83%	0.17%	69.90%	74.14%	0.719
PCAAR method	99.88%	0.12%	73.23%	97.77%	0.837

Filter	Accuracy	Error	Precision	Recall	$F_{1}$
SOR filter	94.37%	5.63%	40.39%	33.15%	0.364
Two-stage SOR filter	93.26%	6.74%	36.48%	52.14%	0.429
ROR filter	95.68%	4.32%	53.41%	86.92%	0.662
AROR filter	96.58%	3.42%	62.22%	75.51%	0.682
Density-based clustering method	96.36%	3.64%	61.38%	66.11%	0.636
PCAAC method	97.25%	2.75%	69.29%	78.15%	0.734
PCAAR method	98.63%	1.37%	78.04%	100%	0.876

Filter	Complexity ( $+$ )	Complexity ( $\times$ )
SOR filter	$O (m^{2} l o g m)$	$O (m^{2} l o g m)$
ROR filter	$O (9 m^{2})$	$O (3 m^{2})$
Density-based clustering method	$O (13 m^{2})$	$O (1.5 m^{2})$
PCAAC method	$O (10 m^{2})$	$O (m^{2})$
PCAAR method	$O (6 m^{2})$	$O (2 m^{2})$

Low-complexity adaptive radius outlier removal filter based on PCA for lidar point cloud denoising

Author Affiliations

Abstract

Cited By

Figures (4)

Tables (6)

Equations (12)

Applied Optics