VTK/ProgrammableMultiVolumeRendering: Difference between revisions

From KitwarePublic
< VTK
Jump to navigationJump to search
(Created page with "! Page under construction ! The goal of this page is to document the shader replacement functionality available in vtkGPUVolumeRaycastMapper starting with version (9?) of VTK.")
 
 
(6 intermediate revisions by the same user not shown)
Line 2: Line 2:


The goal of this page is to document the shader replacement functionality available in vtkGPUVolumeRaycastMapper starting with version (9?) of VTK.
The goal of this page is to document the shader replacement functionality available in vtkGPUVolumeRaycastMapper starting with version (9?) of VTK.
= Fragment shader variables =
== Global variables ==
Variables prefixed with "g_"
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|g_dataPos ||vec3 ||Normalized volume ||Current ray integration position in texture space defined by the first volume, recomputed at every iteration along the ray
|-
|g_terminatePos || vec3 || Normalized volume || Position where we terminate ray integration in texture coordinates of the first volume. This position is computed from the input depth buffer.
|-
|g_eyePosObj ||vec4 || data ||position of the camera in the data space (space of first volume)
|-
|g_fragColor ||vec4 ||None ||Variable where color is accumulated along the ray
|-
|g_srcColor ||vec4 ||None ||Result of the color computation at g_dataPos (current position).
|-
|g_exit ||bool || None ||Terminate ray integration.
|-
|g_skip ||bool || None ||Skip computation of current iteration step if true
|}
== Interpolated variables ==
Variables prefixed with "ip_". These variables contains values computed in the vertex shader stage and interpolated for each fragment.
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|ip_textureCoords ||vec3 || Normalized volume ||Interpolated texture coord
|-
|ip_vertexPos ||vec3 || ? || Interpolated vertex position at current fragment
|-
|ip_inverseTextureDataAdjusted || mat4 || Volume 0 to Texture 0 || Transforms from data space (volume 0) to texture space (volume 0). The matrix is adjusted to take into account the offset between geometry and texture space.
|}
== Uniform variables ==
Variables prefixed with "in_". The value of these variables is set from the c++ code by the mapper. It is now possible for users to define custom uniform variables by using the class vtkOpenGLUniforms. The vtkOpenGLGPURaycastMapper class contains one instance of this class for each type of shader. The objects can be accessed with vtkOpenGLGPURaycastMapper::GetFragmentCustomUniforms and vtkOpenGLGPURaycastMapper::GetVertexCustomUniforms functions. They allow to define uniform variable of any GLSL type and set their values at any moment. These variables can then be used in custom shader code.
=== Input volumes ===
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|in_volume ||sampler3D[n] || None || One sampler for each of the input volumes (n). Allows to sample the 3D texture that contains the volume
|-
|in_volume_scale ||vec4[n] || None || Scale to apply to the sampled scalar value of each input volume
|-
|in_volume_bias ||vec4[n] || None || Offset to apply to the sampled scalar value of each input volume
|-
|in_noOfComponents ||int || None || In the single input volume case, it is possible to feed the mapper with a volume that has multiple components per voxel. This variable contains the number of components.
|-
|in_independentComponents || int ||  None || When a single volume contains multiple components: whether we should treat components as independent. This affects, amongst other things, if a single gradient is computed for all components (dependent) or if one gradient is computed for each component (independent).
|}
=== Other textures ===
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|in_noiseSampler || sampler2D || None || Texture that contains noise and that is used to slightly modify the path of each ray an avoid sampling artefacts.
|-
|in_depthSampler || sampler2D || None || Texture that contains the depth buffer generated by the rendering of polygonal geometry.
|}
=== Camera ===
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|in_cameraPos || vec3 || World || Same as vtkCamera::GetPosition().
|-
|in_projectionMatrix || mat4 || World to Clip || Projection matrix including model view. Transform vertices from world space to clip space.
|-
|in_inverseProjectionMatrix || mat4 || Clip to World || Inverse of projection matrix including model view. Transform vertices from world space to clip space.
|-
| in_modelViewMatrix || mat4 || World to Camera || Transforms vertices from world space to camera space
|-
| in_inverseModelViewMatrix || mat4 || Camera to World || Transforms vertices from camera space to World space
|-
|in_windowLowerLeftCorner || vec2 || screen pixels || Lower left corner of the viewport
|-
|in_inverseOriginalWindowSize || vec2 || screen pixels || 1/(viewport size).
|-
|in_inverseWindowSize || vec2 || screen pixels || 1/(viewport size). 
|}
= Vertex shader variables =
==Uniform variables==
{| class="wikitable"
!Name
!Type
!Space
!Description
|-
|in_cellSpacing || vec3[n] || data? || Distance between voxels for each volume in volume space
|-
|in_modelViewMatrix || mat4 || camera to world || Transforms between camera space and world space
|-
|in_projectionMatrix || mat4 || camera to clip || Projects data from camera space to clip space
|-
|in_volumeMatrix || mat4[n] || data to world || Transforms bounding box of each of the volumes to world coordinates
|-
|in_inverseTextureDatasetMatrix || mat4[n] || data to texture || Transforms from data space to texture space for each volume
|-
|in_cellToPoint || mat4[n] || --- || Corrects for offset between points and OpenGL texture coordinates for each volume
|}

Latest revision as of 20:58, 10 October 2018

! Page under construction !

The goal of this page is to document the shader replacement functionality available in vtkGPUVolumeRaycastMapper starting with version (9?) of VTK.

Fragment shader variables

Global variables

Variables prefixed with "g_"

Name Type Space Description
g_dataPos vec3 Normalized volume Current ray integration position in texture space defined by the first volume, recomputed at every iteration along the ray
g_terminatePos vec3 Normalized volume Position where we terminate ray integration in texture coordinates of the first volume. This position is computed from the input depth buffer.
g_eyePosObj vec4 data position of the camera in the data space (space of first volume)
g_fragColor vec4 None Variable where color is accumulated along the ray
g_srcColor vec4 None Result of the color computation at g_dataPos (current position).
g_exit bool None Terminate ray integration.
g_skip bool None Skip computation of current iteration step if true

Interpolated variables

Variables prefixed with "ip_". These variables contains values computed in the vertex shader stage and interpolated for each fragment.

Name Type Space Description
ip_textureCoords vec3 Normalized volume Interpolated texture coord
ip_vertexPos vec3 ? Interpolated vertex position at current fragment
ip_inverseTextureDataAdjusted mat4 Volume 0 to Texture 0 Transforms from data space (volume 0) to texture space (volume 0). The matrix is adjusted to take into account the offset between geometry and texture space.

Uniform variables

Variables prefixed with "in_". The value of these variables is set from the c++ code by the mapper. It is now possible for users to define custom uniform variables by using the class vtkOpenGLUniforms. The vtkOpenGLGPURaycastMapper class contains one instance of this class for each type of shader. The objects can be accessed with vtkOpenGLGPURaycastMapper::GetFragmentCustomUniforms and vtkOpenGLGPURaycastMapper::GetVertexCustomUniforms functions. They allow to define uniform variable of any GLSL type and set their values at any moment. These variables can then be used in custom shader code.

Input volumes

Name Type Space Description
in_volume sampler3D[n] None One sampler for each of the input volumes (n). Allows to sample the 3D texture that contains the volume
in_volume_scale vec4[n] None Scale to apply to the sampled scalar value of each input volume
in_volume_bias vec4[n] None Offset to apply to the sampled scalar value of each input volume
in_noOfComponents int None In the single input volume case, it is possible to feed the mapper with a volume that has multiple components per voxel. This variable contains the number of components.
in_independentComponents int None When a single volume contains multiple components: whether we should treat components as independent. This affects, amongst other things, if a single gradient is computed for all components (dependent) or if one gradient is computed for each component (independent).

Other textures

Name Type Space Description
in_noiseSampler sampler2D None Texture that contains noise and that is used to slightly modify the path of each ray an avoid sampling artefacts.
in_depthSampler sampler2D None Texture that contains the depth buffer generated by the rendering of polygonal geometry.

Camera

Name Type Space Description
in_cameraPos vec3 World Same as vtkCamera::GetPosition().
in_projectionMatrix mat4 World to Clip Projection matrix including model view. Transform vertices from world space to clip space.
in_inverseProjectionMatrix mat4 Clip to World Inverse of projection matrix including model view. Transform vertices from world space to clip space.
in_modelViewMatrix mat4 World to Camera Transforms vertices from world space to camera space
in_inverseModelViewMatrix mat4 Camera to World Transforms vertices from camera space to World space
in_windowLowerLeftCorner vec2 screen pixels Lower left corner of the viewport
in_inverseOriginalWindowSize vec2 screen pixels 1/(viewport size).
in_inverseWindowSize vec2 screen pixels 1/(viewport size).

Vertex shader variables

Uniform variables

Name Type Space Description
in_cellSpacing vec3[n] data? Distance between voxels for each volume in volume space
in_modelViewMatrix mat4 camera to world Transforms between camera space and world space
in_projectionMatrix mat4 camera to clip Projects data from camera space to clip space
in_volumeMatrix mat4[n] data to world Transforms bounding box of each of the volumes to world coordinates
in_inverseTextureDatasetMatrix mat4[n] data to texture Transforms from data space to texture space for each volume
in_cellToPoint mat4[n] --- Corrects for offset between points and OpenGL texture coordinates for each volume