A novel methodology for the path alignment of visual SLAM in indoor construction inspection
Lu, Tao; Tervola, Sonja; Lü, Xiaoshu; Kibert, Charles J.; Zhang, Qunli; Li, Tong; Yao, Zhitong (2021-07-01)
Lu, Tao
Tervola, Sonja
Lü, Xiaoshu
Kibert, Charles J.
Zhang, Qunli
Li, Tong
Yao, Zhitong
Elsevier
01.07.2021
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2021050328374
https://urn.fi/URN:NBN:fi-fe2021050328374
Kuvaus
vertaisarvioitu
©2021 Elsevier. This manuscript version is made available under the Creative Commons Attribution–NonCommercial–NoDerivatives 4.0 International (CC BY–NC–ND 4.0) license, https://creativecommons.org/licenses/by-nc-nd/4.0/
©2021 Elsevier. This manuscript version is made available under the Creative Commons Attribution–NonCommercial–NoDerivatives 4.0 International (CC BY–NC–ND 4.0) license, https://creativecommons.org/licenses/by-nc-nd/4.0/
Tiivistelmä
Path alignment is the process of mapping an indoor construction inspection path reconstructed by a visual SLAM system onto a 2D map with user interaction required to pinpoint at least two common tie points. In practice, more points are often needed due to path distortions and linear transformations, potentially resulting in reduced productivity. This paper proposes a methodology that combines two novel algorithms for the path alignment: (1) PCA_STAN_ALGO applies principal component analysis to remove path distortions caused by the xz plane of a camera coordinate system not being parallel to the floor plane; and (2) GRPX_TRANS utilizes a graphical user interface to facilitate the path alignment. The proposed methodology enables the users to utilize just two tie points for successful path alignment. An experimental study showed that applying both PCA_STAN_ALGO and GRPX_TRANS saved about 50% in time compared to using only GRPX_TRANS, a result of needing minimal moving points.
Kokoelmat
- Artikkelit [2213]