Upload
others
View
8
Download
0
Embed Size (px)
Citation preview
Computer Graphics & Image Processing Laboratory
Computer Graphics & Image Processing Laboratory320, New Engineering Building 302 School of Computer Science and Engineering, Seoul National University, Seoul, KoreaContact : [email protected]
신영길교수
Publications
Prof. Yeong-Gil Shin (Doctorial Course : 6, Master Course : 4)
[email protected]://cglab.snu.ac.kr/
Education
• Ph.D., Dept of Computer Science, University of Southern California, 1990
• M.S., Dept of Computer Science and Engineering, Seoul National University, 1984
• B.S., Dept of Computer Science and Engineering, Seoul National University, 1981
Experience
• Professor, Computer Graphics & Image Processing Laboratory, 1992 ~ now
• Chief, Information Systems & Technology, Seoul National University, 2014~2016
• Chief, Dept of Computer Science and Engineering, Seoul National University, 2010~2014
• Chief, Institute of computer technology, Seoul National University, 2007~2013
Award history• Excellent engineering projessor award(Industry-Academic Cooperation Part), Seoul
National University, 2019
• The 10th Best Techonology Award, Grand Prize, , 2011
• Korea Software Awards, Ministry of Information and Communication, Republic of Korea,
2002
• Minyoung Chung, Jingyu Lee, Sanguk Park, Chae Eun Lee, Jeongjin Lee, and Yeong-Gil Shin, “Liver Segmentation in Abdominal
CT Images via Auto-Context Neural Network and Self-Supervised Contour Attention”, Artificial Intelligence in Medicine, vol.
113, Mar, 2021.
• Minyoung Chung, Jusang Lee, Sanguk Park, Minkyung Lee, Chae Eun Lee, Jeongjin Lee, and Yeong-Gil Shin, “Individual Tooth
Detection and Identification from Dental Panoramic X-Ray Images via Point-wise Localization and Distance Regularization”,
Artificial Intelligence in Medicine, vol. 111, Jan, 2021.
• Minyoung Chung, Jingyu Lee, Wisoo Song, Youngchan Song, Il-Hyung Yang, Jeongjin Lee, and Yeong-Gil Shin, “Automatic
Registration between Dental Cone-Beam CT and Scanned Surface via Deep-Pose Regression Neural Networks and Clustered
Similarities”, IEEE Transactions on Medical Imaging, vol. 39, Dec, 2020.
• Jiwan Kim, Jeongjin Lee, Minyoung Chung, and Yeong-Gil Shin, “Multiple Weld Seam Extraction from RGB-depth Images for
Automatic Welding via Point Cloud Registration”, Multimedia Tools and Applications, Nov, 2020.
• Minyoung Chung, Jingyu Lee, Minkyung Lee, Jeongjin Lee, and Yeong-Gil Shin, “Deeply Self-Supervised Contour Embedded
Neural Network Applied to Liver Segmentation”, Computer Methods and Programs in Biomedicine, vol. 192, Aug, 2020.
• Minyoung Chung, Minkyung Lee, Jioh Hong, Sanguk Park, Jusang Lee, Jingyu Lee, Il-Hyung Yang, Jeongjin Lee, and Yeong-Gil
Shin, “Pose-Aware Instance Segmentation Framework from Cone-Beam CT Images for Tooth Segmentation”, Computers in
Biology and Medicine, vol. 120, May, 2020.
• Jiseon Kang, Jeongjin Lee, Yeong-Gil Shin, Bohyoung Kim, “Depth-of-Field Rendering Using Progressive Lens Sampling in Direct
Volume Rendering”, IEEE Access 8 (2020): 93335-93345
Computer Graphics & Image Processing Laboratory
We have been focusing on 3D visualization, reconstruction, and image processing of medical images such as CT, MRI, and PET. Recently, we have also expanded our research to industrial CT-based product inspection and defect detection collaborating with many universities and hospitals.Based on these research results, we have commercialized world-class medical image visualization software through cooperation with Infinite, Inc. In addition, we are expanding our activities in the industrial imaging field by developing software for visualization and defect detection.
Computer Graphics & Image Processing Laboratory
Computer Graphics & Image Processing Laboratory320, New Engineering Building 302 School of Computer Science and Engineering, Seoul National University, Seoul, KoreaContact : [email protected]
DEEP LEARNING IN MEDICAL IMAGING
• Object Segmentation Via Deep Convolutional Neural Network (CNN)
Object shape prediction by Edge-to-Contour Neural Network (E2CNet).Image block-wise prediction with CNN classifier.
• Detection of Gynecologic Malignancies in CT images
• Liver Vessel Segmentation and Hepatic/Portal Vein Separation
• Cephalometric Landmark Detection via Deep Convolutional Neural Network (CNN)A system for automatic detection of cephalometric landmarksused for orthodontic diagnosis and treatment planning
Projectional reduction of dimensionality.Graph-based priority queue optimization.
Prof. Yeong-Gil Shin (Doctorial Course : 6, Master Course : 4)
Computer Graphics & Image Processing Laboratory
Prof. Yeong-Gil Shin (Doctorial Course : 6, Master Course : 4)
3D POINTS RECONSTRUCTION FROM 2D IMAGES
𝑖 -th depth Map
𝑃0
𝑃1
𝑃𝑛
𝑃𝑖
Partial point cloud generation from various position via stereo cameraDepth map and point cloud quality enhancement
• Depth map generation
Merging the point cloud sets for whole sceneOptimized 3D point cloud
• 3D point cloud reconstruction
REGISTRATION
• Registration in the integration of laser-scanned mesh into CT images
Surface & CT volume data After registrationSurface & CT volume data After registration
• Real time registration using mesh silhouette
METAL ARTIFACT REDUCTION
Streak artifacts as dark and bright streaks in CTExtract metal segmentation and proper inpainting
Sinogram Reconstructed image
Metal mask imageMetal sinogram
Filtered backprojection
Segmentation
Forward-projection
Correction(interpolation)
Sinogram Reconstructed image
Metal mask imageMetal sinogram
Filtered backprojection
Segmentation
Forward-projection
Correction(interpolation)
Sinogram Reconstructed image
Metal sinogram
Sinogram Reconstructed image
Metal mask imageMetal sinogram
Filtered backprojection
Segmentation
Forward-projection
Correction(interpolation)
Metal mask image
Sinogram Reconstructed image
Metal mask imageMetal sinogram
Filtered backprojection
Segmentation
Forward-projection
Correction(interpolation)
Filtered backprojection
Segmentation
Forward-projection
Correction(interpolation)
Computer Graphics & Image Processing Laboratory320, New Engineering Building 302 School of Computer Science and Engineering, Seoul National University, Seoul, KoreaContact : [email protected]