vtkGreedyTerrainDecimation Class Reference
#include <vtkGreedyTerrainDecimation.h>
Inheritance diagram for vtkGreedyTerrainDecimation:
[legend]Collaboration diagram for vtkGreedyTerrainDecimation:
[legend]List of all members.
Detailed Description
reduce height field (represented as image) to reduced TIN
vtkGreedyTerrainDecimation approximates a height field with a triangle mesh (triangulated irregular network - TIN) using a greedy insertion algorithm similar to that described by Garland and Heckbert in their paper "Fast Polygonal Approximations of Terrain and Height Fields" (Technical Report CMU-CS-95-181). The input to the filter is a height field (represented by a image whose scalar values are height) and the output of the filter is polygonal data consisting of triangles. The number of triangles in the output is reduced in number as compared to a naive tessellation of the input height field. This filter copies point data from the input to the output for those points present in the output.
An brief description of the algorithm is as follows. The algorithm uses a top-down decimation approach that initially represents the height field with two triangles (whose vertices are at the four corners of the image). These two triangles form a Delaunay triangulation. In an iterative fashion, the point in the image with the greatest error (as compared to the original height field) is injected into the triangulation. (Note that the single point with the greatest error per triangle is identified and placed into a priority queue. As the triangulation is modified, the errors from the deleted triangles are removed from the queue, error values from the new triangles are added.) The point whose error is at the top of the queue is added to the triangulaion modifying it using the standard incremental Delaunay point insertion (see vtkDelaunay2D) algorithm. Points are repeatedly inserted until the appropriate (user-specified) error criterion is met.
To use this filter, set the input and specify the error measure to be used. The error measure options are 1) the absolute number of triangles to be produced; 2) a fractional reduction of the mesh (numTris/maxTris) where maxTris is the largest possible number of triangles 2*(dims[0]-1)*(dims[1]-1); 3) an absolute measure on error (maximum difference in height field to reduced TIN); and 4) relative error (the absolute error is normalized by the diagonal of the bounding box of the height field).
- Warning:
- This algorithm requires the entire input dataset to be in memory, hence it may not work for extremely large images. Invoking BoundaryVertexDeletionOff will allow you to stitch together images with matching boundaries.
The input height image is assumed to be positioned in the x-y plane so the scalar value is the z-coordinate, height value.
- See also:
- vtkDecimatePro vtkQuadricDecimation vtkQuadricClustering
- Created by:
-
- CVS contributions (if > 5%):
-
- CVS logs (CVSweb):
/Hybrid/vtkGreedyTerrainDecimation.cxx)/Hybrid/vtkGreedyTerrainDecimation.h)
- Tests:
- vtkGreedyTerrainDecimation (Tests)
Definition at line 99 of file vtkGreedyTerrainDecimation.h.
|
Public Types |
| typedef vtkStructuredPointsToPolyDataFilter | Superclass |
Public Member Functions |
| virtual const char * | GetClassName () |
| virtual int | IsA (const char *type) |
| void | PrintSelf (ostream &os, vtkIndent indent) |
|
| virtual void | SetErrorMeasure (int) |
| virtual int | GetErrorMeasure () |
| void | SetErrorMeasureToNumberOfTriangles () |
| void | SetErrorMeasureToSpecifiedReduction () |
| void | SetErrorMeasureToAbsoluteError () |
| void | SetErrorMeasureToRelativeError () |
|
| virtual void | SetNumberOfTriangles (vtkIdType) |
| virtual vtkIdType | GetNumberOfTriangles () |
|
| virtual void | SetReduction (double) |
| virtual double | GetReduction () |
|
| virtual void | SetAbsoluteError (double) |
| virtual double | GetAbsoluteError () |
|
| virtual void | SetRelativeError (double) |
| virtual double | GetRelativeError () |
|
| virtual void | SetBoundaryVertexDeletion (int) |
| virtual int | GetBoundaryVertexDeletion () |
| virtual void | BoundaryVertexDeletionOn () |
| virtual void | BoundaryVertexDeletionOff () |
Static Public Member Functions |
| int | IsTypeOf (const char *type) |
| vtkGreedyTerrainDecimation * | SafeDownCast (vtkObject *o) |
| vtkGreedyTerrainDecimation * | New () |
Protected Member Functions |
| | vtkGreedyTerrainDecimation () |
| | ~vtkGreedyTerrainDecimation () |
| void | Execute () |
| void | EstimateOutputSize (const vtkIdType numInputPts, vtkIdType &numPts, vtkIdType &numTris) |
| virtual int | SatisfiesErrorMeasure (double error) |
| void | InsertBoundaryVertices () |
| vtkIdType | AddPointToTriangulation (vtkIdType inputPtId) |
| vtkIdType | InsertNextPoint (vtkIdType inputPtId, double x[3]) |
| double * | GetPoint (vtkIdType id) |
| void | GetPoint (vtkIdType id, double x[3]) |
| void | GetTerrainPoint (int i, int j, double x[3]) |
| void | ComputeImageCoordinates (vtkIdType inputPtId, int ij[2]) |
| int | InCircle (double x[3], double x1[3], double x2[3], double x3[3]) |
| vtkIdType | FindTriangle (double x[3], vtkIdType ptIds[3], vtkIdType tri, double tol, vtkIdType nei[3], vtkIdList *neighbors, int &status) |
| void | CheckEdge (vtkIdType ptId, double x[3], vtkIdType p1, vtkIdType p2, vtkIdType tri) |
| void | UpdateTriangles (vtkIdType meshPtId) |
| void | UpdateTriangle (vtkIdType triId, vtkIdType p1, vtkIdType p2, vtkIdType p3) |
| void | UpdateTriangle (vtkIdType triId, int ij1[2], int ij2[2], int ij3[2], double h[4]) |
| int | CharacterizeTriangle (int ij1[2], int ij2[2], int ij[3], int *&min, int *&max, int *&midL, int *&midR, int *&mid, int mid2[2], double h[3], double &hMin, double &hMax, double &hL, double &hR) |
Protected Attributes |
| int | ErrorMeasure |
| vtkIdType | NumberOfTriangles |
| double | Reduction |
| double | AbsoluteError |
| double | RelativeError |
| int | BoundaryVertexDeletion |
| vtkPolyData * | Mesh |
| vtkPointData * | InputPD |
| vtkPointData * | OutputPD |
| vtkDoubleArray * | Points |
| vtkDataArray * | Heights |
| vtkIdType | CurrentPointId |
| double | Tolerance |
| vtkIdList * | Neighbors |
| int | Dimensions [3] |
| double | Origin [3] |
| double | Spacing [3] |
| vtkIdType | MaximumNumberOfTriangles |
| double | Length |
| vtkPriorityQueue * | TerrainError |
| vtkGreedyTerrainDecimationTerrainInfoType * | TerrainInfo |
| vtkGreedyTerrainDecimationPointInfoType * | PointInfo |
Member Typedef Documentation
Constructor & Destructor Documentation
| vtkGreedyTerrainDecimation::vtkGreedyTerrainDecimation |
( |
|
) |
[protected] |
|
Member Function Documentation
| virtual const char* vtkGreedyTerrainDecimation::GetClassName |
( |
|
) |
[virtual] |
|
| int vtkGreedyTerrainDecimation::IsTypeOf |
( |
const char * |
type |
) |
[static] |
|
|
|
Return 1 if this class type is the same type of (or a subclass of) the named class. Returns 0 otherwise. This method works in combination with vtkTypeRevisionMacro found in vtkSetGet.h.
Reimplemented from vtkStructuredPointsToPolyDataFilter. |
| virtual int vtkGreedyTerrainDecimation::IsA |
( |
const char * |
type |
) |
[virtual] |
|
|
|
Return 1 if this class is the same type of (or a subclass of) the named class. Returns 0 otherwise. This method works in combination with vtkTypeRevisionMacro found in vtkSetGet.h.
Reimplemented from vtkStructuredPointsToPolyDataFilter. |
| void vtkGreedyTerrainDecimation::PrintSelf |
( |
ostream & |
os, |
|
|
vtkIndent |
indent |
|
) |
[virtual] |
|
|
|
Methods invoked by print to print information about the object including superclasses. Typically not called by the user (use Print() instead) but used in the hierarchical print process to combine the output of several classes.
Reimplemented from vtkStructuredPointsToPolyDataFilter. |
| virtual void vtkGreedyTerrainDecimation::SetErrorMeasure |
( |
int |
|
) |
[virtual] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image). |
| virtual int vtkGreedyTerrainDecimation::GetErrorMeasure |
( |
|
) |
[virtual] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image). |
| void vtkGreedyTerrainDecimation::SetErrorMeasureToNumberOfTriangles |
( |
|
) |
[inline] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image).
Definition at line 116 of file vtkGreedyTerrainDecimation.h.
References VTK_ERROR_NUMBER_OF_TRIANGLES. |
| void vtkGreedyTerrainDecimation::SetErrorMeasureToSpecifiedReduction |
( |
|
) |
[inline] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image).
Definition at line 118 of file vtkGreedyTerrainDecimation.h.
References VTK_ERROR_SPECIFIED_REDUCTION. |
| void vtkGreedyTerrainDecimation::SetErrorMeasureToAbsoluteError |
( |
|
) |
[inline] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image).
Definition at line 120 of file vtkGreedyTerrainDecimation.h.
References VTK_ERROR_ABSOLUTE. |
| void vtkGreedyTerrainDecimation::SetErrorMeasureToRelativeError |
( |
|
) |
[inline] |
|
|
|
Specify how to terminate the algorithm: either as an absolute number of triangles, a relative number of triangles (normalized by the full resolution mesh), an absolute error (in the height field), or relative error (normalized by the length of the diagonal of the image).
Definition at line 122 of file vtkGreedyTerrainDecimation.h.
References VTK_ERROR_RELATIVE. |
| virtual void vtkGreedyTerrainDecimation::SetNumberOfTriangles |
( |
vtkIdType |
|
) |
[virtual] |
|
|
|
Specify the number of triangles to produce on output. (It is a good idea to make sure this is less than a tessellated mesh at full resolution.) You need to set this value only when the error measure is set to NumberOfTriangles. |
| virtual vtkIdType vtkGreedyTerrainDecimation::GetNumberOfTriangles |
( |
|
) |
[virtual] |
|
|
|
Specify the number of triangles to produce on output. (It is a good idea to make sure this is less than a tessellated mesh at full resolution.) You need to set this value only when the error measure is set to NumberOfTriangles. |
| virtual void vtkGreedyTerrainDecimation::SetReduction |
( |
double |
|
) |
[virtual] |
|
|
|
Specify the reduction of the mesh (represented as a fraction). Note that a value of 0.10 means a 10% reduction. You need to set this value only when the error measure is set to SpecifiedReduction. |
| virtual double vtkGreedyTerrainDecimation::GetReduction |
( |
|
) |
[virtual] |
|
|
|
Specify the reduction of the mesh (represented as a fraction). Note that a value of 0.10 means a 10% reduction. You need to set this value only when the error measure is set to SpecifiedReduction. |
| virtual void vtkGreedyTerrainDecimation::SetAbsoluteError |
( |
double |
|
) |
[virtual] |
|
|
|
Specify the absolute error of the mesh; that is, the error in height between the decimated mesh and the original height field. You need to set this value only when the error measure is set to AbsoluteError. |
| virtual double vtkGreedyTerrainDecimation::GetAbsoluteError |
( |
|
) |
[virtual] |
|
|
|
Specify the absolute error of the mesh; that is, the error in height between the decimated mesh and the original height field. You need to set this value only when the error measure is set to AbsoluteError. |
| virtual void vtkGreedyTerrainDecimation::SetRelativeError |
( |
double |
|
) |
[virtual] |
|
|
|
Specify the relative error of the mesh; that is, the error in height between the decimated mesh and the original height field normalized by the diagonal of the image. You need to set this value only when the error measure is set to RelativeError. |
| virtual double vtkGreedyTerrainDecimation::GetRelativeError |
( |
|
) |
[virtual] |
|
|
|
Specify the relative error of the mesh; that is, the error in height between the decimated mesh and the original height field normalized by the diagonal of the image. You need to set this value only when the error measure is set to RelativeError. |
| virtual void vtkGreedyTerrainDecimation::SetBoundaryVertexDeletion |
( |
int |
|
) |
[virtual] |
|
|
|
Turn on/off the deletion of vertices on the boundary of a mesh. This may limit the maximum reduction that may be achieved. |
| virtual int vtkGreedyTerrainDecimation::GetBoundaryVertexDeletion |
( |
|
) |
[virtual] |
|
|
|
Turn on/off the deletion of vertices on the boundary of a mesh. This may limit the maximum reduction that may be achieved. |
| virtual void vtkGreedyTerrainDecimation::BoundaryVertexDeletionOn |
( |
|
) |
[virtual] |
|
|
|
Turn on/off the deletion of vertices on the boundary of a mesh. This may limit the maximum reduction that may be achieved. |
| virtual void vtkGreedyTerrainDecimation::BoundaryVertexDeletionOff |
( |
|
) |
[virtual] |
|
|
|
Turn on/off the deletion of vertices on the boundary of a mesh. This may limit the maximum reduction that may be achieved. |
| void vtkGreedyTerrainDecimation::Execute |
( |
|
) |
[protected, virtual] |
|
|
|
This method is the old style execute method
Reimplemented from vtkSource. |
| void vtkGreedyTerrainDecimation::EstimateOutputSize |
( |
const vtkIdType |
numInputPts, |
|
|
vtkIdType & |
numPts, |
|
|
vtkIdType & |
numTris |
|
) |
[protected] |
|
| virtual int vtkGreedyTerrainDecimation::SatisfiesErrorMeasure |
( |
double |
error |
) |
[protected, virtual] |
|
| void vtkGreedyTerrainDecimation::InsertBoundaryVertices |
( |
|
) |
[protected] |
|
| vtkIdType vtkGreedyTerrainDecimation::AddPointToTriangulation |
( |
vtkIdType |
inputPtId |
) |
[protected] |
|
| vtkIdType vtkGreedyTerrainDecimation::InsertNextPoint |
( |
vtkIdType |
inputPtId, |
|
|
double |
x[3] |
|
) |
[protected] |
|
| double* vtkGreedyTerrainDecimation::GetPoint |
( |
vtkIdType |
id |
) |
[protected] |
|
| void vtkGreedyTerrainDecimation::GetPoint |
( |
vtkIdType |
id, |
|
|
double |
x[3] |
|
) |
[protected] |
|
| void vtkGreedyTerrainDecimation::GetTerrainPoint |
( |
int |
i, |
|
|
int |
j, |
|
|
double |
x[3] |
|
) |
[protected] |
|
| void vtkGreedyTerrainDecimation::ComputeImageCoordinates |
( |
vtkIdType |
inputPtId, |
|
|
int |
ij[2] |
|
) |
[protected] |
|
| int vtkGreedyTerrainDecimation::InCircle |
( |
double |
x[3], |
|
|
double |
x1[3], |
|
|
double |
x2[3], |
|
|
double |
x3[3] |
|
) |
[protected] |
|
| void vtkGreedyTerrainDecimation::UpdateTriangles |
( |
vtkIdType |
meshPtId |
) |
[protected] |
|
| void vtkGreedyTerrainDecimation::UpdateTriangle |
( |
vtkIdType |
triId, |
|
|
int |
ij1[2], |
|
|
int |
ij2[2], |
|
|
int |
ij3[2], |
|
|
double |
h[4] |
|
) |
[protected] |
|
| int vtkGreedyTerrainDecimation::CharacterizeTriangle |
( |
int |
ij1[2], |
|
|
int |
ij2[2], |
|
|
int |
ij[3], |
|
|
int *& |
min, |
|
|
int *& |
max, |
|
|
int *& |
midL, |
|
|
int *& |
midR, |
|
|
int *& |
mid, |
|
|
int |
mid2[2], |
|
|
double |
h[3], |
|
|
double & |
hMin, |
|
|
double & |
hMax, |
|
|
double & |
hL, |
|
|
double & |
hR |
|
) |
[protected] |
|
Member Data Documentation
The documentation for this class was generated from the following file: