Detailed Description

Provides functions to align a pair of pointclouds using ICP.

The module provides multiple implementation of point-to-plane ICP.

Data Structures
struct	dwICPIterationParams
	Paramter for running ICP between two frames. More...

struct	dwICPParams
	Parameters that control the initialization for ICP module. More...

struct	dwICPResultStats
	Some result stats about the most recent ICP run. More...

Typedefs
typedef dwBool(*	dwICPConvergenceCheck) (const dwTransformation prevSrc2Tgt, const dwTransformation newSrc2Tgt, void *userData)
	Callback function to be executed by ICP module allowing user to overwrite default convergence criteria method. More...

typedef struct dwICPHandle *	dwICPHandle_t

Enumerations
enum	dwICPType { DW_ICP_TYPE_LIDAR_POINT_CLOUD }

Functions
DW_API_PUBLIC dwStatus	dwICP_getCUDAStream (cudaStream_t *stream, dwICPHandle_t icpObj)
	Fetches the ICP module's cuda in a pointer. More...

DW_API_PUBLIC dwStatus	dwICP_getLastResultStats (dwICPResultStats *resultStats, dwICPHandle_t icpObj)
	Get stats about latest ICP run. More...

DW_API_PUBLIC dwStatus	dwICP_initialize (dwICPHandle_t icpObj, const dwICPParams params, dwContextHandle_t ctx)
	Initializes the ICP module. More...

DW_API_PUBLIC dwStatus	dwICP_optimize (dwTransformation outSrc2Tgt, const dwICPIterationParams iParams, dwICPHandle_t icpObj)
	Perform ICP between 2 pointclouds. More...

DW_API_PUBLIC dwStatus	dwICP_release (dwICPHandle_t *icpObj)
	Releases an ICP module. More...

DW_API_PUBLIC dwStatus	dwICP_reset (dwICPHandle_t icpObj)
	Resets the ICP module, clears any data or containers to just-initialized state. More...

DW_API_PUBLIC dwStatus	dwICP_setConvergenceCriteriaCallback (dwICPConvergenceCheck callback, void *userData, dwICPHandle_t icpObj)
	Set callback to be executed to test convergence. More...

DW_API_PUBLIC dwStatus	dwICP_setConvergenceTolerance (float32_t angleTol, float32_t distanceTol, dwICPHandle_t icpObj)
	Set tolerances used by the default ICP convergence criteria method. More...

DW_API_PUBLIC dwStatus	dwICP_setCUDAStream (cudaStream_t stream, dwICPHandle_t icpObj)
	Resets the ICP module's cuda stream to new stream. More...

DW_API_PUBLIC dwStatus	dwICP_setMaxIterations (uint16_t maxIterations, dwICPHandle_t icpObj)
	Set maximum number of iterations which need to be executed. More...

Data Structure Documentation

◆ dwICPIterationParams

struct dwICPIterationParams

Data Fields
const dwTransformation *	initialSource2Target	This column major transform will be applied to the source points before ICP is performed. Final returned transformation will not be pre-multiplied with this
uint64_t	nSourcePts	Sizes of the above arrays, need to be smaller than maxPoints passed via dwICPParams to dwICP_initialize.
uint64_t	nTargetPts
const dwLidarPointXYZI *	sourcePts	Arrays of points, sourcePts need to be transformed by the resulting dwTransformation to align to TargetPts.
const dwLidarPointXYZI *	targetPts

◆ dwICPParams

struct dwICPParams

Data Fields
dwICPType	icpType	Type of the ICP implementation to be used.
uint32_t	maxPoints	Maximum number of points that will be pushed to the ICP optimization. Depending on the implementation this number might be limited (see description of the corresponding method)
cudaStream_t	stream	CudaStream, all the ICP works happens in this stream, can be 0 for default stream.

◆ dwICPResultStats

struct dwICPResultStats

Data Fields
uint32_t	actualNumIterations	How many iterations were actually performed.
float32_t	inlierFraction	Fraction of points which are inliers to the final ICP pose.
float32_t	rmsCost	RMS (weighted) cost after last ICP iteration.

Typedef Documentation

◆ dwICPConvergenceCheck

typedef dwBool(* dwICPConvergenceCheck) (const dwTransformation *prevSrc2Tgt, const dwTransformation *newSrc2Tgt, void *userData)

Callback function to be executed by ICP module allowing user to overwrite default convergence criteria method.

The method will receive original and new transformations as computed by the ICP module

Definition at line 95 of file icp.h.

◆ dwICPHandle_t

typedef struct dwICPHandle* dwICPHandle_t

Definition at line 62 of file icp.h.

Enumeration Type Documentation

◆ dwICPType

enum dwICPType

Enumerator

DW_ICP_TYPE_LIDAR_POINT_CLOUD

ICP implementation optimized for unstructured lidar based point clouds.

A certain corelation of the input points to their index in dwICPIterationParams::[source|target]Pts is expected for good results.

Note: This implementation supports not more than 2^15 points for source and target. If a point cloud is pushed with higher number of points, it will generate a DW_INVALID_ARGUMENT error.

Definition at line 64 of file icp.h.

Function Documentation

◆ dwICP_getCUDAStream()

DW_API_PUBLIC dwStatus dwICP_getCUDAStream	(	cudaStream_t *	stream,
		dwICPHandle_t	icpObj
	)

Fetches the ICP module's cuda in a pointer.

Parameters

[out]	stream	Return cuda stream currently used
[in]	icpObj	The initialized ICP Module to get cuda stream from.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
. DW_INVALID_ARGUMENT - if provided parameters are invalid
DW_SUCCESS

◆ dwICP_getLastResultStats()

DW_API_PUBLIC dwStatus dwICP_getLastResultStats	(	dwICPResultStats *	resultStats,
		dwICPHandle_t	icpObj
	)

Get stats about latest ICP run.

Parameters

[out]	resultStats	Struct with stats about latest ICP run
[in]	icpObj	The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_INVALID_ARGUMENT - if provided parameters are null or invalid
DW_SUCCESS

◆ dwICP_initialize()

DW_API_PUBLIC dwStatus dwICP_initialize	(	dwICPHandle_t *	icpObj,
		const dwICPParams *	params,
		dwContextHandle_t	ctx
	)

Initializes the ICP module.

Initialization will perform any required memory allocations on GPU and CPU.

Parameters

[out]	icpObj	A pointer to the ICP handle for the created module.
[in]	params	Parameters for initilizations.
[in]	ctx	Specifies the handler to the context under which the ICP module is created.

Returns: DW_BAD_CAST - if the provided context handle is invalid.
DW_INVALID_ARGUMENT - if provided parameters are invalid
DW_SUCCESS

◆ dwICP_optimize()

DW_API_PUBLIC dwStatus dwICP_optimize	(	dwTransformation *	outSrc2Tgt,
		const dwICPIterationParams *	iParams,
		dwICPHandle_t	icpObj
	)

Perform ICP between 2 pointclouds.

The method will return a transformation which aligns source points to target points.

Parameters

[out]	outSrc2Tgt	Given trtransformation will be overwritten with the ICP result.
[in]	iParams	Pointer to iteration parameters.
[in]	icpObj	The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided context or ICP handle is invalid.
DW_INVALID_ARGUMENT - if provided parameters are invalid
DW_SUCCESS

Note: Result returned in outSrc2Tgt will be premultiplied with the initial transformation passed in iParams. Final equation to align source to target is then targetPt = outSrc2Tgt * sourcePt

◆ dwICP_release()

DW_API_PUBLIC dwStatus dwICP_release ( dwICPHandle_t * icpObj )

Releases an ICP module.

Parameters

[in] icpObj The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_INVALID_ARGUMENT - if provided parameters are invalid
DW_SUCCESS

◆ dwICP_reset()

DW_API_PUBLIC dwStatus dwICP_reset ( dwICPHandle_t icpObj )

Resets the ICP module, clears any data or containers to just-initialized state.

Parameters

[in] icpObj The initialized ICP Module to reset.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_SUCCESS

◆ dwICP_setConvergenceCriteriaCallback()

DW_API_PUBLIC dwStatus dwICP_setConvergenceCriteriaCallback	(	dwICPConvergenceCheck	callback,
		void *	userData,
		dwICPHandle_t	icpObj
	)

Set callback to be executed to test convergence.

If nullptr, the default test will be used.

Parameters

[in]	callback	Method to execute to test for convergence
[in]	userData	User data to be passed to the convergence callback
[in]	icpObj	The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_SUCCESS

◆ dwICP_setConvergenceTolerance()

DW_API_PUBLIC dwStatus dwICP_setConvergenceTolerance	(	float32_t	angleTol,
		float32_t	distanceTol,
		dwICPHandle_t	icpObj
	)

Set tolerances used by the default ICP convergence criteria method.

Parameters

[in]	angleTol	Change in radians between two consecutive iteration steps
[in]	distanceTol	Change in units between two consecutive iteration steps
[in]	icpObj	The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_SUCCESS

Note: both tolerances has to be fulfilled, to declare convergence (i.e. it is a binary AND).

◆ dwICP_setCUDAStream()

DW_API_PUBLIC dwStatus dwICP_setCUDAStream	(	cudaStream_t	stream,
		dwICPHandle_t	icpObj
	)

Resets the ICP module's cuda stream to new stream.

Parameters

[in]	stream	Cuda stream to be used by ICP implementation.
[in]	icpObj	The initialized ICP module to set stream for.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
. DW_INVALID_ARGUMENT - if provided parameters are invalid
DW_SUCCESS

◆ dwICP_setMaxIterations()

DW_API_PUBLIC dwStatus dwICP_setMaxIterations	(	uint16_t	maxIterations,
		dwICPHandle_t	icpObj
	)

Set maximum number of iterations which need to be executed.

Note that ICP might converge earlier, due to the tolerances set in dwICP_setConvergenceTolerance().

Parameters

[in]	maxIterations	Maximal number of iterations to execute.
[in]	icpObj	The initialized ICP Module.

Returns: DW_BAD_CAST - if the provided ICP handle is invalid.
DW_SUCCESS

DriveWorks SDK Reference

| 0.6.67 Release

Detailed Description

Data Structures

Typedefs

Enumerations

Functions

Data Structure Documentation

◆ dwICPIterationParams

◆ dwICPParams

◆ dwICPResultStats

Typedef Documentation

◆ dwICPConvergenceCheck

◆ dwICPHandle_t

Enumeration Type Documentation

◆ dwICPType

Function Documentation

◆ dwICP_getCUDAStream()

◆ dwICP_getLastResultStats()

◆ dwICP_initialize()

◆ dwICP_optimize()

◆ dwICP_release()

◆ dwICP_reset()

◆ dwICP_setConvergenceCriteriaCallback()

◆ dwICP_setConvergenceTolerance()

◆ dwICP_setCUDAStream()

◆ dwICP_setMaxIterations()