File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1111/mice.12714
- Scopus: eid_2-s2.0-85124053074
- WOS: WOS:000667243200001
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Indoor camera pose estimation via style‐transfer 3D models
Title | Indoor camera pose estimation via style‐transfer 3D models |
---|---|
Authors | |
Issue Date | 2022 |
Citation | Computer-Aided Civil and Infrastructure Engineering, 2022, v. 37 n. 3, p. 335-353 How to Cite? |
Abstract | Many vision-based indoor localization methods require tedious and comprehensive pre-mapping of built environments. This research proposes a mapping-free approach to estimating indoor camera poses based on a 3D style-transferred building information model (BIM) and photogrammetry technique. To address the cross-domain gap between virtual 3D models and real-life photographs, a CycleGAN model was developed to transform BIM renderings into photorealistic images. A photogrammetry-based algorithm was developed to estimate camera pose using the visual and spatial information extracted from the style-transferred BIM. The experiments demonstrated the efficacy of CycleGAN in bridging the cross-domain gap, which significantly improved performance in terms of image retrieval and feature correspondence detection. With the 3D coordinates retrieved from BIM, the proposed method can achieve near real-time camera pose estimation with an accuracy of 1.38 m and 10.1° in indoor environments. |
Persistent Identifier | http://hdl.handle.net/10722/301961 |
ISSN | 2023 Impact Factor: 8.5 2023 SCImago Journal Rankings: 2.972 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Chen, J | - |
dc.contributor.author | Li, S | - |
dc.contributor.author | Liu, DH | - |
dc.contributor.author | Lu, WW | - |
dc.date.accessioned | 2021-08-21T03:29:32Z | - |
dc.date.available | 2021-08-21T03:29:32Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Computer-Aided Civil and Infrastructure Engineering, 2022, v. 37 n. 3, p. 335-353 | - |
dc.identifier.issn | 1093-9687 | - |
dc.identifier.uri | http://hdl.handle.net/10722/301961 | - |
dc.description.abstract | Many vision-based indoor localization methods require tedious and comprehensive pre-mapping of built environments. This research proposes a mapping-free approach to estimating indoor camera poses based on a 3D style-transferred building information model (BIM) and photogrammetry technique. To address the cross-domain gap between virtual 3D models and real-life photographs, a CycleGAN model was developed to transform BIM renderings into photorealistic images. A photogrammetry-based algorithm was developed to estimate camera pose using the visual and spatial information extracted from the style-transferred BIM. The experiments demonstrated the efficacy of CycleGAN in bridging the cross-domain gap, which significantly improved performance in terms of image retrieval and feature correspondence detection. With the 3D coordinates retrieved from BIM, the proposed method can achieve near real-time camera pose estimation with an accuracy of 1.38 m and 10.1° in indoor environments. | - |
dc.language | eng | - |
dc.relation.ispartof | Computer-Aided Civil and Infrastructure Engineering | - |
dc.title | Indoor camera pose estimation via style‐transfer 3D models | - |
dc.type | Article | - |
dc.identifier.email | Chen, J: chenjj10@hku.hk | - |
dc.identifier.email | Lu, WW: wilsonlu@hku.hk | - |
dc.identifier.authority | Lu, WW=rp01362 | - |
dc.identifier.doi | 10.1111/mice.12714 | - |
dc.identifier.scopus | eid_2-s2.0-85124053074 | - |
dc.identifier.hkuros | 324444 | - |
dc.identifier.volume | 37 | - |
dc.identifier.issue | 3 | - |
dc.identifier.spage | 335 | - |
dc.identifier.epage | 353 | - |
dc.identifier.isi | WOS:000667243200001 | - |