Download presentation
Presentation is loading. Please wait.
Published byRandolf Sutton Modified over 9 years ago
1
Volume Graphics (graduate course) Bong-Soo Sohn School of Computer Science and Engineering Chung-Ang University
2
Course Overview Level : CSE graduate course No Textbook –We will use lecture notes, recent papers, and several handouts. Lecture Format –Lectures by Instructor (half) + Student Presentation (half) –Project Result Presentation
3
Topics Volume Rover (Volume Visualization SW)
4
Main Topics Volume Rover 에서 사용되는 Volume Visualization 기술 –Volume Representation –Volume Rendering, Transfer Function Design, MIP –Isosurface Extraction, Contour Spectrum, Contour Tree –(Multicore and Manycore) Parallel Processing –CUDA isosurface extraction –Salient Contour Extraction –Point/Image Based Geometric Processing, Meshing (Voronoi Diagram, Delaunay Triangulation) –Shape Analysis –3D Image Processing (Smoothing)
5
Course Information Time : Mon 3pm-6pm Place : 208-529 Instructor Information –Office : 208-523-2 –email : bongbong@cau.ac.krbongbong@cau.ac.kr –Office Tel : 820-5843 –Office Hour : individual email appointment
6
Image and Geometric Processing 3D/4D Image CT/MRI Electron Microscopy OCT Simulation Geometric Modeling Processing Filtering, Segmentation Visualization Quantification (Structure Analysis) Laser Scanner Point Cloud OBJECTOBJECT Engineering Research Scientific Research Biomedical Research Building/Plant Construction
7
Input Biomedical Images Rapid Advance of Imaging Techniques Multiscale Static(3D) vs time-varying(4D) Molecular Level (Angstrom Scale) Cellular and Tissue Level (Nano Scale) Organ Level (Micro Scale) Organ Level Cryo-EM Electron Microscopy OCT (Optical Coherence Tomography) CT/MRI X-ray Crystallography
8
Building Information Modeling (BIM) generation and management of a digital representation of physical and functional characteristics of a facility.
9
Salient Feature Analysis Salient Contour Extraction –Useful for segmentation, analysis and visualization of regions of interest –Can be applied to CAD(Computer Aided Diagnosis) for detecting suspicious regions 9 mass (tumor) dense tissue breast boundarypectoral muscle
10
Contour Tree
11
KISTI 수퍼컴퓨팅센터
12
Cardiovascular Modeling Research Pipeline 3D Image Acquisition Geometric Modeling Simulation Rendering, Quantitative Visualization cardivascular disease research, medical device design, and surgical planning
13
Sulcal Morphology Analysis (courtesy of Dr. J.-K. Seong) Reduced average sulcal curvature and depth in AD (Im et al. NeuroImage 2008)
14
Biomedical OCT Visualization OCT(Optical Coherence Tomography) Non-invasive optical tomographic imaging technique with micrometer scale resolution. Widely accepted in biomedical applications Contribution Real-time volume visualization of 3-dimensional OCT images. ( Journal of Korean Physical Society [SCI], 2007 ) 3D Volume Visualization
15
Lecture Schedule Visualization Overview (1 week) Scalar Visualization Techniques (4 weeks) –Volume Visualization Basics, Volume Representation –Volume Rendering Ray casting, HW accelerated volume rendering MIP (Maximum Intensity Projection) Transfer function design –Isocontour Visualization Marching Cubes + Accelerated method Quantitative and Topological Analysis (Contour Tree) Large Data Visualization (parallelism, out-of-core, compression) Interactive Visualization Interface –Illustrative Visualization, NPR in Visualization
16
Lecture Schedule Vector Visualization Techniques (1 week) –Line Integral Convolution, Streamline Image Based Geometric Modeling (1 weeks) –Filtering –Segmentation (Level Set Method) –Mesh Generation Shape Analysis (2 weeks) –Voronoi Diagram, Delaunay Triangulation –Medial Axis Algorithms, Skeletonization –Shape Matching, Salient Feature Extraction –Surface Property (curvature, …) –Applications (Surface Reconstruction, Protein Docking, …)
17
Volume Rendering, Isocontour 3D World is modeled with a function (= image) F(x,y,z) (e.g. CT : human body density) Surface is modeled with a level set of a function level set = isosurface = isocontour = implicit surface { (x,y,z) | F(x,y,z) = w } ( w is a fixed value, called isovalue ) Level set may represent important features of a function e.g. skin surface ( w =skin density) or bone surface ( w =bone density) in body CT
18
Example (Volume Rendering, Isocontour) [ volume image ] [ skin surface ] [ bone surface ] F(x,y,z) Level Set : F(x,y,z) = w w = skin density w = bone density
19
Hybrid Parallel Contour Extraction Different from isocontour extraction Divide contour extraction process into –Propagation Iterative algorithm -> hard to optimize using GPU multi-threaded algorithm executed in multi-core CPU –Triangulation CUDA implementation executed in many-core GPU 19
20
Interactive Interface with Quantitative Information Geometric Property as saliency level –Gradient(color) + Area (thickness) 20
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.