Difference between revisions of "Slicer-IGT:GPU-IGT:01/09/08 T-con with Ben and Nicholas at 3pm Salt Lake City Time"

From NAMIC Wiki
Jump to: navigation, search
Line 26: Line 26:
 
# some debuggings of registration CUDA codes
 
# some debuggings of registration CUDA codes
 
# start to work on itk classes:
 
# start to work on itk classes:
## itkResampleImageFilter
+
## itkCudaResampleImageFilter
## itkImageToImageMetric
+
## itkCudaImageToImageMetric
  
 
==Misc==
 
==Misc==

Revision as of 22:41, 9 January 2008

Home < Slicer-IGT:GPU-IGT:01 < 09 < 08 T-con with Ben and Nicholas at 3pm Salt Lake City Time

Topics

Slicer Integration

Cuda Support Classes

  1. Implement the host access side such as allocate memory, delete and so forth in vtk-Classes
  2. Combining effort in:
    1. vtkCudaMemory
    2. vtkCudaAlgorithm
    3. vtkCuda
    4. vtkTexture

CUDA algorithms

Volume Rendering

Important features:

  1. use Camera Coordinates and Projection Matrix instead of rotationMatrix
    1. Currently planar projection
  2. use Z-Buffer from previous render
    1. easy todo.
  3. Lighting Model use multiple lights
  4. Color gradients rather than 2 colors

Volume Rendering CUDA module

Move all the code from the cuda Code into a Support Classes

Registration

  1. some debuggings of registration CUDA codes
  2. start to work on itk classes:
    1. itkCudaResampleImageFilter
    2. itkCudaImageToImageMetric

Misc

Build

  1. CUDA on Cygwin: use the --ccbin flag to set the compiler bin directory. (failed for me)

Results

CUDA COLOR

COlor Map for to write to cuda

Resources

  1. Slicer-IGT/GPU-IGT
  2. Volume Rendering With Cuda