Module Type & Category
Authors, Collaborators & Contact
- Andreas Freudling, Student Intern at SPL (freudling at bwh.harvard.edu)
- Alex Yarmarkovich, Isomics, SPL
- GPU accelerated raycasting by Yanling Liu, National Cancer Institute
- Contact: Alex Yarmarkovich, alexy at bwh.harvard.edu
Volume Rendering allows the rendering of volumes in 3D view in semi transparent mode, potentially combined with the models in one scene.
- Use "Load and Save" panel to select a volume to render, select predefined parameter sets and save your custom settings.
- Use "Performance" panel to select different rendering techniques (mappers): CPU Ray Casting, GPU Ray Casting, and OpenGL Polygon Blending (texture mapping). By default CPU Ray Casting is used for maximum compatibility. In all three methods, image quality is automatically determined based on user expected frame rate. For example, if user requested 20 fps using the Expected Frame Rate slider, all three mappers will try to decrease resolution or sampling rate to reach 20 frames per second (0.05 second rendering time per frame). If user requested 1fps, then all three mappers will try to increase resolution or sampling rate for better quality since longer rendering time is available (1 second rendering time per frame). For each mapper, a Force High Quality check button is provided to override frame rate input and to generate high quality images (low performance).
- Use "Threshold" panel to adjust transfer functions that map scalar values at each voxel to opacity and color.
- Use "Cropping" panel to define a region of interest in your volume.
- "Advanced" panel allows you to fine tune the transfer functions that map scalar values to opacity and color.
Examples, Use Cases & Tutorials
- Tutorials on usage of Volume Rendering Module (with movies!) are available here.
Quick Tour of Features and Use
- Expected Frame Rate Slider: the lower the requested fps (frame per second), the longer the rendering time is available for better quality image. When requesting higher number of fps, mappers will have shorter available rendering time so that image quality is reduced for expected performance. The default value is 5 fps for a balance between image quality and performance.
- CPU Ray Casting: two rendering modes: compositing and maximum intensity projection (MIP). Check the Maximum Intensity Projection check button to enable MIP rendering.
- GPU Ray Casting: by default lighting is off since some low end graphics card may have trouble to handle lighting in GPU ray casting. Internally the mapper has maximum size limitation of input volume (256x256x256) due to limited graphics memory resource. Input volume larger than 256x256x256 will be resampled to 256x256x256. When large graphics memory is available, for example 1G memory on GTX 280, user may check the "Large Volume Size" check button to increase size limit to 512x512x512. However, small size volume will not be super sampled. MIP rendering is also available in GPU ray casting. Check the Maximum Intensity Projection check button to enable MIP rendering.
- Stereo rendering: the GPU ray casting mapper supports active glasses stereo rendering at interactive frame rate.
Tested Hardware Configuration
- Nvidia Quadro 5600,
- Nvidia Quadro 1700,
- Nvidia GTX 280,
- Nvidia Quadro 3700M,
- NVIDIA Quadro NVS 140M
- Nvidia Quadro 4600,
- Nvidia GeForce 8600,
- AMD/ATI Mobility Radeon X1400 (lighting not supported),
- Intermixing polygon models and GPU ray casting may lead to incorrect depth information.
Follow this link to the Slicer3 bug tracker.
Follow this link to the Slicer3 bug tracker. Please select the usability issue category when browsing or contributing.
Follow this link to source code in ViewVC.
Documentation generated by doxygen.