Point Cloud

Cloud of 3D points. More...

Functions
cvbres_t	CVC3DCreateDensePointCloud (cvbdim_t Width, cvbdim_t Height, cvbval_t Flags, CVCOMPOSITE &PointCloud)
	Creates a new dense Cartesian 3D PointCloud. More...
cvbres_t	CVC3DCreateSparsePointCloud (cvbint64_t NumPoints, cvbval_t Flags, CVCOMPOSITE &PointCloud)
	Creates a new sparse Cartesian 3D PointCloud. More...
cvbres_t	CVC3DPointCloudAnalyzeLayout (CVCOMPOSITE PointCloud, CVC3DPointCloudLayout &Layout)
	Analyzes the available buffers/planes of the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetDatatype (CVCOMPOSITE PointCloud, cvbdatatype_t &DT)
	Gets the cvbdatatype_t of the x,y,z(,w) components of the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetLatticeSize (CVCOMPOSITE PointCloud, cvbdim_t &Width, cvbdim_t &Height)
	Gets the number of x,y,z(,w) point rows and columns of the given dense PointCloud. More...
cvbres_t	CVC3DPointCloudGetNumPoints (CVCOMPOSITE PointCloud, cvbint64_t &NumPoints)
	Gets the number of x,y,z(,w) points in the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointCDsPointer (CVCOMPOSITE PointCloud, CVC3DPointCD *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointCFsPointer (CVCOMPOSITE PointCloud, CVC3DPointCF *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointComponentsPointers (CVCOMPOSITE PointCloud, void *&pXs, intptr_t &XsInc, void *&pYs, intptr_t &YsInc, void *&pZs, intptr_t &ZsInc, void *&pWs, intptr_t &WsInc, size_t &NumPoints)
	Tries to get the point components from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointDsPointer (CVCOMPOSITE PointCloud, CVC3DPointD *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointFsPointer (CVCOMPOSITE PointCloud, CVC3DPointF *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointHDsPointer (CVCOMPOSITE PointCloud, CVC3DPointHD *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointHFsPointer (CVCOMPOSITE PointCloud, CVC3DPointHF *&pPoints, size_t &NumPoints)
	Tries to get the 3D points from the given PointCloud. More...
cvbres_t	CVC3DPointCloudGetPointsConfidencePointer (CVCOMPOSITE PointCloud, void *&pConfidences, intptr_t &Increment, cvbdatatype_t &DataType)
	Tries to get the confidence components from the given PointCloud. More...

Detailed Description

Cloud of 3D points.

At its core the PointCloud is a collection of 3D points (x,y,z, and optionally w). The components' data type can either be float or double. To be flexible for various use cases additional information can also be aggregated via the IComposite.

A point cloud is a CVCOMPOSITE with at least three planes denoting Cartesian point components. If possible, these point components are aggregated in one or more __IPFNCBuffer__s. Other information like an ITransformation can be attached to the IComposite.

The functions in this section simplify PointCloud access, so you don't need to enumerate and analyze the **IPlane**s manually. Here you find functions to access the points (or point components), the number of points and the structural information like cvbdatatype_t and CVC3DPointCloudLayout.

Sparse and Dense **PointCloud**s

Sparse **PointCloud**s are – logically seen, not from a memory perspective – an array of points. The IPlane objects aggregated by the cloud's IComposite are one-dimensional (rank of 1). They give no assurance of the ordering of the points.

Dense **PointCloud**s are placed on a lattice with a given width and height. Their **IPlane**s are two-dimensional (rank of 2). The advantage of dense clouds is, that you have immediate neighbor information. The disadvantage is, that you cannot have arbitrary clouds (e.g. rotating a dense cloud quickly breaks the lattice view). Also points behind other points cannot be addressed (same x,y components with differing z components).

In memory there are no gaps in a dense cloud. You need to work around that with a confidence IPlane (CVCPR_PixConfidence plane role). And as dense clouds are also just points, you can access dense clouds like sparse clouds.

The easiest way to determine whether you have a dense point cloud is to call CVC3DPointCloudGetLatticeSize.

Confidence Plane

Confidence planes are especially important to dense **PointCloud**s as we cannot remove points without destroying the lattice. The confidence for a point ranges from 0 to 1 (inclusive), where 0 is 0% and 1 is 100%. A value of 0.5 thus has 50% confidence.

A confidence value in the confidence plane relates to the point with the same index in the x, y, z (, w) component planes. There is also a special point type available (CVC3DPointCF/CVC3DPointCD) which makes the confidence information directly available. You can create such **PointCloud**s by adding CVC3DPCF_XYZConfidence to the creation flags.

Memory Layout

__PointCloud__s do not have a fixed buffer layout. Following Common Vision Blox's plane model, each point component can be enumerated individually. With this approach various buffer layouts can be handled (see CVC3DPointCloudLayout for additional information on the possible layouts).

You can enumerate these planes and buffers with the CVCore library's IPlaneEnum, IPlane, IComposite and IPFNCBuffer functions.

Function Documentation

CVC3DCreateDensePointCloud()

cvbres_t CVC3DCreateDensePointCloud	(	cvbdim_t	Width,
		cvbdim_t	Height,
		cvbval_t	Flags,
		CVCOMPOSITE &	PointCloud
	)

Creates a new dense Cartesian 3D PointCloud.

The resulting point cloud will always have a contiguous point array (CVC3DPCL_Interleaved) with the cvbdatatype_t of the point components being DT_Float | DT_Signed | 32 if CVC3DPCF_DTFloat is set in the Flags; DT_Float | DT_Signed | 64 if CVC3DPCF_DTDouble is set. The buffer's PFNC format is Coord3D_ABC32f if CVC3DPCF_XYZ is set; CVC3DPCF_XYZW and CVC3DPCF_XYZConfidence have no PFNC representation.

Creating the point cloud with with CVC3DPCF_XYZConfidence will provide interleaved XYZConfidence (PointCloudLayout will be CVC3DPCL_Linear), whereas creating it with CVC3DPCF_WithConfidence, only XYZ will be interleaved but not the confidence plane (PointCloudLayout will be CVC3DPCL_Interleaved).

Note

The content of the created point cloud is only allocated, but not guaranteed to be initialized with any value.

When the CVC3DPCF_XYZConfidence flag is used the confidence plane is also interleaved (see CVC3DPointCF/CVC3DPointCD) and has the data type of the point components.

If this function returns successfully you need to release the returned PointCloud via ReleaseObject if not needed anymore.

Parameters

[in]	Width	Number of Cartesian X, Y, Z(, W) points on the x-component of the lattice.
[in]	Height	Number of Cartesian X, Y, Z(, W) points on the y-component of the lattice.
[in]	Flags	CVC3DPointCloudFlags specifying the kind of point cloud to be created.
[out]	PointCloud	Variable to receive the handle of the point cloud object as a CVCOMPOSITE.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_PARAMETER) if NumPoints is less or equal to 0 or Flags have an unsupported combination.
• CVC_ERROR (CVC_E_MEMORY) if not enough memory was available to create the point cloud.

See also

PointCloud, CVC3DPointCloudAnalyzeLayout, CVC3DPointCloudGetDatatype, CVC3DPointCloudGetNumPoints, CVC3DPointCloudGetPointFsPointer, CVC3DPointCloudGetPointDsPointer, CVC3DPointCloudGetPointHFsPointer, CVC3DPointCloudGetPointHDsPointer, CVC3DCreateDensePointCloud, CVC3DPointCloudGetPointCFsPointer, CVC3DPointCloudGetPointCDsPointer

CVC3DCreateSparsePointCloud()

cvbres_t CVC3DCreateSparsePointCloud	(	cvbint64_t	NumPoints,
		cvbval_t	Flags,
		CVCOMPOSITE &	PointCloud
	)

Creates a new sparse Cartesian 3D PointCloud.

The resulting point cloud will have a contiguous point array (CVC3DPCL_Interleaved) with the cvbdatatype_t of the point components being DT_Float | DT_Signed | 32 if CVC3DPCF_DTFloat is set in the Flags; DT_Float | DT_Signed | 64 if CVC3DPCF_DTDouble is set. The buffer's PFNC format is Coord3D_ABC32f if CVC3DPCF_XYZ is set; CVC3DPCF_XYZW and CVC3DPCF_XYZConfidence have no PFNC representation.

Creating the point cloud with with CVC3DPCF_XYZConfidence will provide interleaved XYZConfidence (PointCloudLayout will be CVC3DPCL_Linear), whereas creating it with CVC3DPCF_WithConfidence, only XYZ will be interleaved but not the confidence plane (PointCloudLayout will be CVC3DPCL_Interleaved).

Note

The content of the created point cloud is only allocated, but not guaranteed to be initialized with any value.

If this function returns successfully you need to release the returned PointCloud via ReleaseObject if not needed anymore.

Parameters

[in]	NumPoints	Number of Cartesian X, Y, Z(, W) points.
[in]	Flags	CVC3DPointCloudFlags specifying the kind of point cloud to be created.
[out]	PointCloud	Variable to receive the handle of the point cloud object as a CVCOMPOSITE.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_PARAMETER) if NumPoints is less or equal to 0 or Flags have an unsupported combination.
• CVC_ERROR (CVC_E_MEMORY) if not enough memory was available to create the point cloud.

See also

PointCloud, CVC3DPointCloudAnalyzeLayout, CVC3DPointCloudGetDatatype, CVC3DPointCloudGetNumPoints, CVC3DPointCloudGetPointFsPointer, CVC3DPointCloudGetPointDsPointer, CVC3DPointCloudGetPointHFsPointer, CVC3DPointCloudGetPointHDsPointer, CVC3DCreateDensePointCloud

CVC3DPointCloudAnalyzeLayout()

cvbres_t CVC3DPointCloudAnalyzeLayout	(	CVCOMPOSITE	PointCloud,
		CVC3DPointCloudLayout &	Layout
	)

Analyzes the available buffers/planes of the given PointCloud.

This function queries the IPFNCBuffer and then the IPlaneEnum interfaces.

Parameters

[in]	PointCloud	Point cloud to analyze.
[out]	Layout	CVC3DPointCloudLayout of PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the point components in a known, compatible layout.

See also

CVC3DPointCloudLayout for additional layout information.

CVC3DPointCloudGetDatatype()

cvbres_t CVC3DPointCloudGetDatatype	(	CVCOMPOSITE	PointCloud,
		cvbdatatype_t &	DT
	)

Gets the cvbdatatype_t of the x,y,z(,w) components of the given PointCloud.

Note

This function does more than just to query the IPlaneEnum or IPFNCBuffer interfaces. It also checks whether PointCloud will be treated as a point cloud.

Parameters

[in]	PointCloud	Point cloud of which the point component cvbdatatype_t is to queried.
[out]	DT	The cvbdatatype_t of the x,y,z(,w) components of the PointCloud points.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the point components in a known, compatible layout.

CVC3DPointCloudGetLatticeSize()

cvbres_t CVC3DPointCloudGetLatticeSize	(	CVCOMPOSITE	PointCloud,
		cvbdim_t &	Width,
		cvbdim_t &	Height
	)

Gets the number of x,y,z(,w) point rows and columns of the given dense PointCloud.

Note

This function does more than just to query the IPlaneEnum or IPFNCBuffer interfaces. It also checks whether PointCloud will be treated as a dense point cloud.

Parameters

[in]	PointCloud	Point cloud of which the number of points is to be queried.
[out]	Width	Variable to be filled with the number of columns.
[out]	Height	Variable to be filled with the number of rows.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_INVALIDDIMENSION) if PointCloud is not a dense point cloud.
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the point components in a known, compatible layout.

See also

CVC3DPointCloudGetNumPoints

CVC3DPointCloudGetNumPoints()

cvbres_t CVC3DPointCloudGetNumPoints	(	CVCOMPOSITE	PointCloud,
		cvbint64_t &	NumPoints
	)

Gets the number of x,y,z(,w) points in the given PointCloud.

Note

One a dense point cloud this is equivalent to this code:

cvbint64_t GetNumPoints(CVCOMPOSITE PointCloud)
{
  cvbdim_t width = 0, height = 0;
  if(cvbres_t error = CVC3DPointCloudGetLatticeSize(PointCloud, width, height))
    throw std::logic_error("not a dense point cloud");
  
  return static_cast<cvbint64_t>(width) * height;
}

This function does more than just to query the IPlaneEnum or IPFNCBuffer interfaces. It also checks whether PointCloud will be treated as a point cloud.

Parameters

[in]	PointCloud	Point cloud of which the number of points is to be queried.
[out]	NumPoints	Variable to be filled with the number of points in the PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the point components in a known, compatible layout.

See also

CVC3DPointCloudGetLatticeSize

CVC3DPointCloudGetPointCDsPointer()

cvbres_t CVC3DPointCloudGetPointCDsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointCD *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z, C float points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointCD structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointCFsPointer()

cvbres_t CVC3DPointCloudGetPointCFsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointCF *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z, C float points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointCF structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointComponentsPointers()

cvbres_t CVC3DPointCloudGetPointComponentsPointers	(	CVCOMPOSITE	PointCloud,
		void *&	pXs,
		intptr_t &	XsInc,
		void *&	pYs,
		intptr_t &	YsInc,
		void *&	pZs,
		intptr_t &	ZsInc,
		void *&	pWs,
		intptr_t &	WsInc,
		size_t &	NumPoints
	)

Tries to get the point components from the given PointCloud.

This function gets the pointers to the internal buffer(s) of X, Y, Z point components.

Note

This function supports all linear Cartesian point clouds. If the component planes are stored interleaved as e.g. CVC3DPointF, you can also use the CVC3DPointCloudGetPointFsPointer function to access the points as an array.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points components are to be accessed.
[out]	pXs	Variable to receive the pointer to the first X-component of the first point.
[out]	XsInc	Increment to the next X-component of the next point in bytes.
[out]	pYs	Variable to receive the pointer to the first Y-component of the first point.
[out]	YsInc	Increment to the next Y-component of the next point in bytes.
[out]	pZs	Variable to receive the pointer to the first Z-component of the first point.
[out]	ZsInc	Increment to the next Z-component of the next point in bytes.
[out]	pWs	Variable to receive the pointer to the first W-component of the first point (if present; nullptr if not).
[out]	WsInc	Increment to the next W-component of the next point in bytes (if present; 0 if not).
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the point components in a known, compatible layout.

See also

CVC3DPointCloudGetPointFsPointer, CVC3DPointCloudAnalyzeLayout, CVC3DPointCloudGetPointsConfidencePointer

Example

#include <limits>
#include <CVCError.h>
#include <iCVCore3D.h>
 
struct Range
{
  float Min;
  float Max;
};
 
cvbres_t CalculateBoundingBox(CVCOMPOSITE PointCloud, Range &XRange, Range &YRange, Range &ZRange)
{
  cvbdatatype_t dt = 0;
  if(cvbres_t error = CVC3DPointCloudGetDatatype(PointCloud, dt))
    return error;
  if(BytesPerPixel(dt) != sizeof(float)) // very simple check
    return CVC_ERROR(CVC_E_INVALIDDATATYPE); // only support float-clouds
 
  intptr_t Xs = 0, XsInc = 0;
  intptr_t Ys = 0, YsInc = 0;
  intptr_t Zs = 0, ZsInc = 0;
  intptr_t Ws = 0, WsInc = 0; 
  size_t numPoints = 0;
  if(cvbres_t error = CVC3DPointCloudGetPointComponentsPointers(PointCloud, reinterpret_cast<void *&>(Xs), XsInc, 
                                                                            reinterpret_cast<void *&>(Ys), YsInc, 
                                                                            reinterpret_cast<void *&>(Zs), ZsInc, 
                                                                            reinterpret_cast<void *&>(Ws), WsInc, 
                                                                numPoints))
    return error;
 
  if(Ws != 0)
    return CVC_ERROR(CVC_E_NOTSUPPORTED); // only x,y,z clouds
  
  XRange.Min = YRange.Min = ZRange.Min = std::numeric_limits<float>::max();
  XRange.Max = YRange.Max = ZRange.Max = std::numeric_limits<float>::lowest();
  
  for(size_t pointIndex = 0; pointIndex < numPoints; ++pointIndex)
  {
    float x = *reinterpret_cast<float*>(Xs + pointIndex * XsInc);
    float y = *reinterpret_cast<float*>(Ys + pointIndex * YsInc);
    float z = *reinterpret_cast<float*>(Zs + pointIndex * ZsInc);
    
    if(x < XRange.Min)
      XRange.Min = x;
    if(x > XRange.Max)
      XRange.Max = x;
    if(y < YRange.Min)
      YRange.Min = y;
    if(y > YRange.Max)
      YRange.Max = y;
    if(z < ZRange.Min)
      ZRange.Min = Z;
    if(z > ZRange.Max)
      ZRange.Max = z;
  }
  
  return CVC_ERROR(CVC_E_OK);
}

CVC3DPointCloudGetPointDsPointer()

cvbres_t CVC3DPointCloudGetPointDsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointD *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z double points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointD structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointFsPointer()

cvbres_t CVC3DPointCloudGetPointFsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointF *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z float points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointF structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointHDsPointer()

cvbres_t CVC3DPointCloudGetPointHDsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointHD *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z, W double points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointHD structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointHFsPointer()

cvbres_t CVC3DPointCloudGetPointHFsPointer	(	CVCOMPOSITE	PointCloud,
		CVC3DPointHF *&	pPoints,
		size_t &	NumPoints
	)

Tries to get the 3D points from the given PointCloud.

This function gets the pointer to the internal array of X, Y, Z, W float points if it available.

Note

This function only succeeds if the component planes are stored interleaved in the PointCloud. For all other Cartesian point clouds you can use CVC3DPointCloudGetPointComponentsPointers.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points are to be accessed.
[out]	pPoints	Variable to be filled with pointer to first point in PointCloud.
[out]	NumPoints	Variable to be filled with the number of points in PointCloud.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not store the points as contiguous CVC3DPointHF structs.

See also

CVC3DPointCloudGetPointComponentsPointers, CVC3DPointCloudAnalyzeLayout

CVC3DPointCloudGetPointsConfidencePointer()

cvbres_t CVC3DPointCloudGetPointsConfidencePointer	(	CVCOMPOSITE	PointCloud,
		void *&	pConfidences,
		intptr_t &	Increment,
		cvbdatatype_t &	DataType
	)

Tries to get the confidence components from the given PointCloud.

This is a companion function to be used with any point access function like CVC3DPointCloudGetPointComponentsPointers to access the internal buffer of the confidence plane.

Note

It is possible that the confidence plane has a different cvbdatatype_t then the point data.

If the point cloud is also dense, then you can access the other rows by using the rowIncrement = pointIncrement * latticeWidth (see CVC3DPointCloudGetLatticeSize)

Parameters

[in]	PointCloud	Point cloud of which the points confidence are to be accessed.
[out]	pConfidences	Variable to receive the pointer to the first confidence component of the first point.
[out]	Increment	Increment to the next confidence component of the next point in bytes.
[out]	DataType	Variable to receive the data type of the confidence plane.

Returns

• CVC_ERROR (CVC_E_OK) on success
• CVC_ERROR (CVC_E_WRONGOBJECT) if PointCloud is not a valid point cloud IComposite.
• CVC_ERROR (CVC_E_INVALIDDATATYPE) if PointCloud does not contain a confidence plane in a known form.

See also

CVC3DPointCloudGetPointComponentsPointers