Undersampled non-uniform density multi-station 3D point cloud alignment method based on manifold clustering

Jixiang NIE; Yibo WANG; Qiubing SHEN; Heping HUANG; Xiaolin CHEN; Hui CHEN

doi:10.37188/CJLCD.2023-0318

Chinese Journal of Liquid Crystals and Displays, Volume. 39, Issue 9, 1255(2024)

Undersampled non-uniform density multi-station 3D point cloud alignment method based on manifold clustering

Jixiang NIE¹, Yibo WANG², Qiubing SHEN¹, Heping HUANG³, Xiaolin CHEN⁴, and Hui CHEN^1、*

Author Affiliations

¹College of Automation Engineering，Shanghai University of Electric Power，Shanghai 200090，China

²West-East Gas Transmission Branch，National Petroleum Gas Pipeline Group Co. Ltd.，Shanghai 200122，China

³Zhengtai Instrument（Hangzhou）Co. Ltd.，Hangzhou 310052，China

⁴Shanghai Chint Power System Co. Ltd.，Shanghai 201600，China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(12)

Fig. 1. Flow chart of undersampling method based on manifold clustering

Download full size

View in Article

Fig. 2. Two-dimensional manifold surface

Download full size

View in Article

Fig. 3. Segmentation method of non-uniform density point cloud

Download full size

View in Article

Fig. 4. Flow chart of the sampling process

Download full size

View in Article

Fig. 5. Comparison of chamfer distances after refinement for different data sets

Download full size

View in Article

Fig. 6. Visualization comparison of sampling effects of different methods for three datasets

Download full size

View in Article

Fig. 7. Comparison of visualisation of the alignment effect of different methods on the Bunny dataset

Download full size

View in Article

Table 1. Point cloud data set information
View table
View in Article
Table 1. Point cloud data set information
Dataset Bunny Dragon Happy
Number of views 10 15 15
Total points 362 272 469 193 1 099 005

Table 2. Sampling results of Bunny dataset
View table
View in Article
Table 2. Sampling results of Bunny dataset
Method Pre-sampling points Post-sampling points Time/s Chamfer distance/mm
Random sampling method 362 272 10 000 0.054 0.185
Uniform sampling method 362 272 9 995 0.205 0.041
Proposed method 362 272 10 002 3.952 0.065

Table 3. Sampling results of Dragon dataset
View table
View in Article
Table 3. Sampling results of Dragon dataset
Method Pre-sampling points Post-sampling points Time/s Chamfer distance/mm
Random sampling method 469 193 12 000 0.065 0.216
Uniform sampling method 469 193 12 008 0.150 0.045
Proposed method 469 193 12 006 4.922 0.071

Table 4. Sampling results of Happy dataset

View table

View in Article

Table 4. Sampling results of Happy dataset

Method	Pre-sampling points	Post-sampling points	Time/s	Chamfer distance/mm
Random sampling method	1 099 005	12 000	0.120	0.235
Uniform sampling method	1 099 005	11 972	0.246	0.068
Proposed method	1 099 005	12 017	5.431	0.084

Table 5. Accuracy comparison of different methods
View table
View in Article
Table 5. Accuracy comparison of different methods
Method E_R/mm E_t/mm Time/s
Initial 0.040 9 3.350 3 -
K-means 0.019 9 1.310 7 5.677 2
Random sampling method 0.033 9 2.045 2 1.383 8
Proposed method 0.018 4 1.584 7 1.318 5

Tools

Get Citation

Copy Citation Text

Jixiang NIE, Yibo WANG, Qiubing SHEN, Heping HUANG, Xiaolin CHEN, Hui CHEN. Undersampled non-uniform density multi-station 3D point cloud alignment method based on manifold clustering[J]. Chinese Journal of Liquid Crystals and Displays, 2024, 39(9): 1255

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites