3D skin surface reconstruction from a single image by merging global curvature and local texture using the guided filtering for 3D haptic palpation. Issue 4 (11th May 2018)
- Record Type:
- Journal Article
- Title:
- 3D skin surface reconstruction from a single image by merging global curvature and local texture using the guided filtering for 3D haptic palpation. Issue 4 (11th May 2018)
- Main Title:
- 3D skin surface reconstruction from a single image by merging global curvature and local texture using the guided filtering for 3D haptic palpation
- Authors:
- Lee, K.
Kim, M.
Kim, K. - Abstract:
- Abstract: Background: Skin surface evaluation has been studied using various imaging techniques. However, all these studies had limited impact because they were performed using visual exam only. To improve on this scenario with haptic feedback, we propose 3D reconstruction of the skin surface using a single image. Unlike extant 3D skin surface reconstruction algorithms, we utilize the local texture and global curvature regions, combining the results for reconstruction. Materials and Methods: The first entails the reconstruction of global curvature, achieved by bilateral filtering that removes noise on the surface while maintaining the edge (ie, furrow) to obtain the overall curvature. The second entails the reconstruction of local texture, representing the fine wrinkles of the skin, using an advanced form of bilateral filtering. The final image is then composed by merging the two reconstructed images. Results: We tested the curvature reconstruction part by comparing the resulting curvatures with measured values from real phantom objects while local texture reconstruction was verified by measuring skin surface roughness. Then, we showed the reconstructed result of our proposed algorithm via the reconstruction of various real skin surfaces. The experimental results demonstrate that our approach is a promising technology to reconstruct an accurate skin surface with a single skin image. Conclusion: We proposed 3D skin surface reconstruction using only a single camera. WeAbstract: Background: Skin surface evaluation has been studied using various imaging techniques. However, all these studies had limited impact because they were performed using visual exam only. To improve on this scenario with haptic feedback, we propose 3D reconstruction of the skin surface using a single image. Unlike extant 3D skin surface reconstruction algorithms, we utilize the local texture and global curvature regions, combining the results for reconstruction. Materials and Methods: The first entails the reconstruction of global curvature, achieved by bilateral filtering that removes noise on the surface while maintaining the edge (ie, furrow) to obtain the overall curvature. The second entails the reconstruction of local texture, representing the fine wrinkles of the skin, using an advanced form of bilateral filtering. The final image is then composed by merging the two reconstructed images. Results: We tested the curvature reconstruction part by comparing the resulting curvatures with measured values from real phantom objects while local texture reconstruction was verified by measuring skin surface roughness. Then, we showed the reconstructed result of our proposed algorithm via the reconstruction of various real skin surfaces. The experimental results demonstrate that our approach is a promising technology to reconstruct an accurate skin surface with a single skin image. Conclusion: We proposed 3D skin surface reconstruction using only a single camera. We highlighted the utility of global curvature, which has not been considered important in the past. Thus, we proposed a new method for 3D reconstruction that can be used for 3D haptic palpation, dividing the concepts of local and global regions. … (more)
- Is Part Of:
- Skin research and technology. Volume 24:Issue 4(2018)
- Journal:
- Skin research and technology
- Issue:
- Volume 24:Issue 4(2018)
- Issue Display:
- Volume 24, Issue 4 (2018)
- Year:
- 2018
- Volume:
- 24
- Issue:
- 4
- Issue Sort Value:
- 2018-0024-0004-0000
- Page Start:
- 672
- Page End:
- 685
- Publication Date:
- 2018-05-11
- Subjects:
- 3D reconstruction with single image -- image guided filtering -- skin curvature reconstruction -- skin texture reconstruction
Skin -- Research -- Periodicals
Skin -- Diseases -- Periodicals
Skin -- Physiology -- Periodicals
616.5 - Journal URLs:
- http://www.blackwellpublishing.com/journal.asp?ref=0909-752X&site=1 ↗
http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1600-0846 ↗
http://onlinelibrary.wiley.com/ ↗ - DOI:
- 10.1111/srt.12584 ↗
- Languages:
- English
- ISSNs:
- 0909-752X
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 8295.948000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 7970.xml