Indoor Structure Method Based on Motion Assisted by Geomagnetic Features

Zhoumeng He; Guoliang Chen; Mingcong Shu; Kaiyu Di; Hu Liu

doi:10.3788/LOP232682

Laser & Optoelectronics Progress, Volume. 61, Issue 18, 1837012(2024)

Indoor Structure Method Based on Motion Assisted by Geomagnetic Features

Zhoumeng He^1,2, Guoliang Chen^1,2、*, Mingcong Shu^1,2, Kaiyu Di^1,2, and Hu Liu^1,2

¹School of Environment and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, Jiangsu, China

²Key Laboratory of Land Environment and Disaster Monitoring of Ministry of Natural Resources, China University of Mining and Technology, Xuzhou 221116, Jiangsu, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(10)

Fig. 1. Workflow of indoor structure from motion assisted by geomagnetic features

Download full size

Fig. 2. Process of matching points between local models

Download full size

Fig. 3. Schematic diagrams of experimental scenes and location relationship of data acquisition devices. (a) Experimental scene on the same floor; (b) experimental scene on different floors; (c) device location relationship

Download full size

Fig. 4. Data preprocessing. (a) Smoothing of geomagnetic sequences; (b) original image (left) and smooth image (right)

Download full size

Fig. 5. Comparison of reconstruction effect of proposed method and traditional method. (a) Proposed method; (b) traditional method

Download full size

Table 1. Collected data and preliminary processed relevant information
View table
Table 1. Collected data and preliminary processed relevant information
Dataset Before processing After processing
Number of geomagnetisms Number of images Number of geomagnetisms Number of images
Same floor 33000 19800 991 991
Different floors 27400 16440 1662 1662

Table 2. Accuracy table of three algorithms for clustering geomagnetic data
View table
Table 2. Accuracy table of three algorithms for clustering geomagnetic data
Algorithm K-Means Bi-KMeans DBSCAN
Precision 56.36 89.09 70.91

Table 3. The local reconstruction of proposed method does not run RGPA and the reconstruction of different data subsets by traditional method
View table
Table 3. The local reconstruction of proposed method does not run RGPA and the reconstruction of different data subsets by traditional method
Dataset Proposed method Dataset Traditional method
a a
b a+b
c a+b+c
d a+b+c+d
e a+b+c+d+e

Table 4. Comparison reconstructions of proposed method and traditional method
View table
Table 4. Comparison reconstructions of proposed method and traditional method
Reconstructed view Proposed method Traditional method
Front view
Left view
Top view
Complete model

Table 5. Comparison performance of traditional method and proposed method

View table

Table 5. Comparison performance of traditional method and proposed method

Dataset	Number of images	Traditional method					Proposed method
Dataset	Number of images	Time /s	Camera registration rate /%	Sparse point clouds /10³	Point cloud generation rate /s	Single-image time-consuming /s	Time /s	Camera registration rate /%	Sparse point clouds /10³	Point cloud generation rate /s^-1	Single-image time-consuming /s
Same floor	991	1906	84	148.001	78	1.92	975	100	158.876	163	0.98
Different floors	1662	3986	81	261.622	66	2.40	2960	85	274.541	93	1.78

Tools

Get Citation

Copy Citation Text

Zhoumeng He, Guoliang Chen, Mingcong Shu, Kaiyu Di, Hu Liu. Indoor Structure Method Based on Motion Assisted by Geomagnetic Features[J]. Laser & Optoelectronics Progress, 2024, 61(18): 1837012

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites