Zhan Yu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Zhan Yu has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Jingyi Yu
ShanghaiTech University, Shanghai, China
3 shared publications Karl V Steiner
University of Delaware
2 shared publicationsKenneth E Barner
University of Delaware
2 shared publications- Thomas L. Bauer
Department of Thoracic Surgery, Christiana Care Health System (CCHS), PO Box 1668, Wilmington, DE 19899, USA; Helen F. Graham Cancer Center and Research Institute, Christiana Care Health System (CCHS), PO Box 1668, Wilmington, DE 19899, USA; The Value Institute, Christiana Care Health System (CCHS), PO Box 1668, Wilmington, DE 19899, USA
2 shared publications - Xinqing Guo
University of Delaware Newark DE USA
2 shared publications Andrew Lumsdaine
Indiana Univ. (United States)
1 shared publication
Latest Publications
- 1 January 2014
- Vol. 196, 479-85
An immersive surgery training system with live streaming capability.
Providing real-time, interactive immersive surgical training has been a key research area in telemedicine. Earlier approaches have mainly adopted videotaped training that can only show imagery from a fixed view point. Recent advances on commodity 3D imaging have enabled a new paradigm for immersive surgical training by acquiring nearly complete 3D reconstructions of actual surgical procedures. However, unlike 2D videotaping that can easily stream data in real-time, by far 3D imaging based solutions require pre-capturing and processing the data; surgical trainings using the data have to be conducted offline after the acquisition. In this paper, we present a new real-time immersive 3D surgical training system. Our solution builds upon the recent multi-Kinect based surgical training system [1] that can acquire and display high delity 3D surgical procedures using only a small number of Microsoft Kinect sensors. We build on top of the system a client-server model for real-time streaming. On the server front, we efficiently fuse multiple Kinect data acquired from different viewpoints and compress and then stream the data to the client. On the client front, we build an interactive space-time navigator to allow remote users (e.g., trainees) to witness the surgical procedure in real-time as if they were present in the room.
- 1 January 2014
- Vol. 196, 479-85
An immersive surgery training system with live streaming capability.
Providing real-time, interactive immersive surgical training has been a key research area in telemedicine. Earlier approaches have mainly adopted videotaped training that can only show imagery from a fixed view point. Recent advances on commodity 3D imaging have enabled a new paradigm for immersive surgical training by acquiring nearly complete 3D reconstructions of actual surgical procedures. However, unlike 2D videotaping that can easily stream data in real-time, by far 3D imaging based solutions require pre-capturing and processing the data; surgical trainings using the data have to be conducted offline after the acquisition. In this paper, we present a new real-time immersive 3D surgical training system. Our solution builds upon the recent multi-Kinect based surgical training system [1] that can acquire and display high delity 3D surgical procedures using only a small number of Microsoft Kinect sensors. We build on top of the system a client-server model for real-time streaming. On the server front, we efficiently fuse multiple Kinect data acquired from different viewpoints and compress and then stream the data to the client. On the client front, we build an interactive space-time navigator to allow remote users (e.g., trainees) to witness the surgical procedure in real-time as if they were present in the room.
- 7 March 2013
- Published by SPIE-Intl Soc Optical Eng
- Vol. 8667
Plenoptic depth map in the case of occlusions
Recent realizations of hand-held plenoptic cameras have given rise to previously unexplored effects in photography. Designing a mobile phone plenoptic camera is becoming feasible with the significant increase of computing power of mobile devices and the introduction of System on a Chip. However, capturing high numbers of views is still impractical due to special requirements such as ultra-thin camera and low costs. In this paper, we analyze a mobile plenoptic camera solution with a small number of views. Such a camera can produce a refocusable high resolution final image if a depth map is generated for every pixel in the sparse set of views. With the captured multi-view images, the obstacle to recovering a high-resolution depth is occlusions. To robustly resolve these, we first analyze the behavior of pixels in such situations. We show that even under severe occlusion, one can still distinguish different depth layers based on statistics. We estimate the depth of each pixel by discretizing the space in the scene and conducting plane sweeping. Specifically, for each given depth, we gather all corresponding pixels from other views and model the in-focus pixels as a Gaussian distribution. We show how it is possible to distinguish occlusion pixels, and in-focus pixels in order to find the depths. Final depth maps are computed in real scenes captured by a mobile plenoptic camera.
- 7 March 2013
- Published by SPIE-Intl Soc Optical Eng
- Vol. 8667
Plenoptic depth map in the case of occlusions
Recent realizations of hand-held plenoptic cameras have given rise to previously unexplored effects in photography. Designing a mobile phone plenoptic camera is becoming feasible with the significant increase of computing power of mobile devices and the introduction of System on a Chip. However, capturing high numbers of views is still impractical due to special requirements such as ultra-thin camera and low costs. In this paper, we analyze a mobile plenoptic camera solution with a small number of views. Such a camera can produce a refocusable high resolution final image if a depth map is generated for every pixel in the sparse set of views. With the captured multi-view images, the obstacle to recovering a high-resolution depth is occlusions. To robustly resolve these, we first analyze the behavior of pixels in such situations. We show that even under severe occlusion, one can still distinguish different depth layers based on statistics. We estimate the depth of each pixel by discretizing the space in the scene and conducting plane sweeping. Specifically, for each given depth, we gather all corresponding pixels from other views and model the in-focus pixels as a Gaussian distribution. We show how it is possible to distinguish occlusion pixels, and in-focus pixels in order to find the depths. Final depth maps are computed in real scenes captured by a mobile plenoptic camera.
- 1 January 2013
- Vol. 184, 161-7
A portable immersive surgery training system using RGB-D sensors.
Surgical training plays an important role in assisting residents to develop critical skills. Providing effective surgical training, however, remains as a challenging task. Existing videotaped training instructions can only show imagery from a fixed viewpoint that lacks both depth perception and interactivity. We present a new portable immersive surgical training system that is capable of acquiring and displaying high fidelity 3D reconstructions of actual surgical procedures. Our solution utilizes a set of Microsoft Kinect sensors to simultaneously recover the participants, the surgical environment, and the surgical scene itself. We then develop a space-time navigator to allow the trainees to witness and explore a prior procedure as if they were there. Preliminary feedback from residents shows that our system is much more effective than conventional videotaped system.
- 1 January 2013
- Vol. 184, 161-7
A portable immersive surgery training system using RGB-D sensors.
Surgical training plays an important role in assisting residents to develop critical skills. Providing effective surgical training, however, remains as a challenging task. Existing videotaped training instructions can only show imagery from a fixed viewpoint that lacks both depth perception and interactivity. We present a new portable immersive surgical training system that is capable of acquiring and displaying high fidelity 3D reconstructions of actual surgical procedures. Our solution utilizes a set of Microsoft Kinect sensors to simultaneously recover the participants, the surgical environment, and the surgical scene itself. We then develop a space-time navigator to allow the trainees to witness and explore a prior procedure as if they were there. Preliminary feedback from residents shows that our system is much more effective than conventional videotaped system.