Functions for Match3D DNC (CAD-based 3D-object recognition). More...

Data Structures
struct	CVDNCResult
	Definition of a specific detection result. More...

struct	CVDNCSearchParams
	Definition of search parameters. More...

struct	CVDNCTeachParams
	Definition of teach parameters which are used to create a DNC-classifier from a CVDNCSAMPLES object. More...

Typedefs
typedef void *	CVDNCFINDER
	Definition of DNC-classifier object. More...

typedef void *	CVDNCRESULTS
	Definition of DNC-results object. More...

typedef cvbbool_t(*	CVDNCTeachCallback) (cvbint32_t Done, cvbint32_t Total, void *pPrivate)
	Callback definition for teach-process. More...

Functions
cvbres_t	CVDNCCreateFromFile (const char *FileName, CVDNCFINDER &Dncf)
	Create an instance of DNC-classifier from file (ASCII string version). More...

cvbres_t	CVDNCCreateFromFileW (const wchar_t *FileName, CVDNCFINDER &Dncf)
	Create an instance of DNC-classifier from file (UNICODE string version). More...

cvbres_t	CVDNCCreateFromSamples (CVDNCSAMPLES Dncs, CVDNCTeachParams Params, CVDNCFINDER &Dncf, CVDNCTeachCallback CallbackFunc, void *pPrivate)
	Create a DNC-classifier from an instance of DNC-samples. More...

cvbres_t	CVDNCFind (CVDNCFINDER Dncf, CVCOMPOSITE DenseCloud, CVDNCRESULTS &Results)
	Perform a search task. More...

cvbint32_t	CVDNCGetNumResults (CVDNCRESULTS Results)
	Read size of result list. More...

cvbres_t	CVDNCGetResult (CVDNCRESULTS Results, cvbdim_t Index, CVDNCResult &Result)
	Get values for an individual search result. More...

cvbres_t	CVDNCGetSearchParams (CVDNCFINDER Dncf, CVDNCSearchParams &params)
	Retrieve search parameters. More...

cvbres_t	CVDNCGetTeachParams (CVDNCFINDER Dncf, double &Resolution, double &Fringe, CVDNCTeachParams &Params)
	Retrieve the teach parameters of the classifier. More...

cvbres_t	CVDNCSaveClassifier (CVDNCFINDER Dncf, const char *FileName)
	Write current classifier to file (ASCII string version). More...

cvbres_t	CVDNCSaveClassifierW (CVDNCFINDER Dncf, const wchar_t *FileName)
	Write current classifier to file (UNICODE string version). More...

cvbres_t	CVDNCSetSearchParams (CVDNCFINDER Dncf, CVDNCSearchParams params)
	Specify search parameters. More...

Detailed Description

Functions for Match3D DNC (CAD-based 3D-object recognition).

Note, that this library allows only finding 3D objects using a given DNC finder.The DNC finder must be created via the CVB Teachbench.exe first.The Teachbench.exe can be found in cvb%/Applications.

Typedef Documentation

◆ CVDNCFINDER

CVDNCFINDER

Definition of DNC-classifier object.

There are various functions to create a DNC-classifier object, see CVDNCCreateFromSamples, CVDNCCreateFromFile, CVDNCCreateFromFileW. Each of these functions return a pointer of type CVDNCFINDER, which is used in subsequent function calls to address this object.

When the object is no longer used, it must be released with a call to ReleaseObject.

See also: CVDNCCreateFromSamples, CVDNCCreateFromFile, CVDNCCreateFromFileW

◆ CVDNCRESULTS

CVDNCRESULTS

Definition of DNC-results object.

A (successful) call of CVDNCFind returns a pointer to a DNC-results object. This object represents a list of individual results. Special functions give access to these results, e.g. CVDNCGetNumResults and CVDNCGetResult.

When the object is no longer used, it must be released with a call to ReleaseObject.

See also: CVDNCFind, CVDNCGetNumResults, CVDNCGetResult, CVDNCResult

◆ CVDNCTeachCallback

CVDNCTeachCallback

Callback definition for teach-process.

During teaching (see CVDNCCreateFromSamples) a callback will be executed repeatedly, indicating progress. At each call, the number of samples to teach (Total) and the number of already teached samples (Done) will be given. The first call (initialization) gives 0 for both Done and Total.

You can control the termination of the teaching process by setting the return value of this callback to false. In this case, teaching will terminate and CVDNCCreateFromSamples will return an error.

Parameters

[in]	Done	Amount of progress (0 for initialization, 1..Total otherwise).
[in]	Total	Number of samples to teach. (0 for initialization, >0 otherwise).
[in]	pPrivate	Address of callback's owner object.

Returns

Teaching will be stopped if return value is false.
Teaching will continue if return value is true.

See also: CVDNCCreateFromSamples

Function Documentation

◆ CVDNCCreateFromFile()

cvbres_t CVDNCCreateFromFile	(	const char *	FileName,
		CVDNCFINDER &	Dncf
	)

Create an instance of DNC-classifier from file (ASCII string version).

This function loads a previously stored DNC-classifier file and returns a pointer to a CVDNCFINDER object.

Parameters

[in]	FileName	Path to classifier file.
[out]	Dncf	Pointer to DNC-classifier object.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCCreateFromFileW

◆ CVDNCCreateFromFileW()

cvbres_t CVDNCCreateFromFileW	(	const wchar_t *	FileName,
		CVDNCFINDER &	Dncf
	)

Create an instance of DNC-classifier from file (UNICODE string version).

This function loads a previously stored DNC classifier file and returns a pointer to a CVDNCFINDER object.

Parameters

[in]	FileName	Path to classifier file.
[out]	Dncf	Pointer to DNC-classifier object.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCCreateFromFile

◆ CVDNCCreateFromSamples()

cvbres_t CVDNCCreateFromSamples	(	CVDNCSAMPLES	Dncs,
		CVDNCTeachParams	Params,
		CVDNCFINDER &	Dncf,
		CVDNCTeachCallback	CallbackFunc,
		void *	pPrivate
	)

Create a DNC-classifier from an instance of DNC-samples.

This function converts a set of examples created from a CAD model into a CVDNCFINDER object that can be used to search for the model in a point cloud. The examples are given in the form of the CVDNCSAMPLES object Dncs.

Classifier creation is controlled by the teach parameters Params and the callback function CallbackFunc. If CallbackFunc is set and returns false, the classifier creation is terminated immediately.

If you do not want or need a callback mechanism, you can set CallbackFunc and pPrivate to Null.

Parameters

[in]	Dncs	Pointer to existing DNC-samples object.
[in]	Params	Teach parameters.
[out]	Dncf	Pointer to newly created DNC-classifier object.
[in]	CallbackFunc	Teach callback function.
[in]	pPrivate	Pointer to CallbackFunc's owner.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncs is not a CVDNCSAMPLES object
CVC_ERROR(CVC_E_PARAMETER) if some of the values in Params are out of range
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCTeachParams, CVDNCTeachCallback

◆ CVDNCFind()

cvbres_t CVDNCFind	(	CVDNCFINDER	Dncf,
		CVCOMPOSITE	DenseCloud,
		CVDNCRESULTS &	Results
	)

Perform a search task.

Detect objects in the given point cloud and return a list of results. Try to adjust the search parameters if there are too few results or the result list contains false positives.

The result list must be released after use by calling ReleaseObject.

The point cloud DenseCloud must be of dense format, otherwise an error is returned. It must also either be orientated in sensor frame coordinates, or the point cloud must be extrinsically calibrated such that the direction of the Z-axis points to the sensor.

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[in]	DenseCloud	The point cloud in which objects are to be detected.
[out]	Results	Pointer to result list.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_WRONGOBJECT) if DenseCloud is not a dense COMPOSITE
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCGetNumResults, CVDNCGetResult, CVDNCSetSearchParams

◆ CVDNCGetNumResults()

cvbint32_t CVDNCGetNumResults ( CVDNCRESULTS Results )

Read size of result list.

This function returns the number of CVDNCResult-entries contained in the given result list Results.

Parameters

[in] Results Pointer to result list.

Returns

Number of entries in result list on success
CVC_ERROR(CVC_E_WRONGOBJECT) else

See also: CVDNCGetResult

◆ CVDNCGetResult()

cvbres_t CVDNCGetResult	(	CVDNCRESULTS	Results,
		cvbdim_t	Index,
		CVDNCResult &	Result
	)

Get values for an individual search result.

Retrieve specific information of an individual result in Results.

Parameters

[in]	Results	Pointer to result list.
[in]	Index	Index of result entry.
[out]	Result	A CVDNCResult structure filled with the corresponding data.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_PARAMETER) if Index is out of range
CVC_ERROR(CVC_E_WRONGOBJECT) if Results is not a CVDNCRESULTS-object

See also: CVDNCGetNumResults

◆ CVDNCGetSearchParams()

cvbres_t CVDNCGetSearchParams	(	CVDNCFINDER	Dncf,
		CVDNCSearchParams &	params
	)

Retrieve search parameters.

This function returns the current search parameters.

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[out]	params	Search parameters.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCSetSearchParams

◆ CVDNCGetTeachParams()

cvbres_t CVDNCGetTeachParams	(	CVDNCFINDER	Dncf,
		double &	Resolution,
		double &	Fringe,
		CVDNCTeachParams &	Params
	)

Retrieve the teach parameters of the classifier.

This function returns the parameters with which the classifier was trained. The values for Resolution and Fringe are not teach parameters in the strict sense - they rather concern the sample generation itself -, but they are given for the sake of completeness.

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[out]	Resolution	Sampling resolution in pixels per millimeter.
[out]	Fringe	Fringe around object during sampling in mm.
[out]	Params	Teach parameters.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_ERROR) else

See also: DNCSSetResolution, DNCSSetFringe CVDNCCreateFromSamples

◆ CVDNCSaveClassifier()

cvbres_t CVDNCSaveClassifier	(	CVDNCFINDER	Dncf,
		const char *	FileName
	)

Write current classifier to file (ASCII string version).

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[in]	FileName	Path to classifier file.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCSaveClassifierW

◆ CVDNCSaveClassifierW()

cvbres_t CVDNCSaveClassifierW	(	CVDNCFINDER	Dncf,
		const wchar_t *	FileName
	)

Write current classifier to file (UNICODE string version).

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[in]	FileName	Path to classifier file.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCSaveClassifier

◆ CVDNCSetSearchParams()

cvbres_t CVDNCSetSearchParams	(	CVDNCFINDER	Dncf,
		CVDNCSearchParams	params
	)

Specify search parameters.

This function specifies the search parameters for subsequent calls of CVDNCFind.

Parameters

[in]	Dncf	Pointer to DNC-classifier object.
[in]	params	Search parameters.

Returns

CVC_ERROR(CVC_E_OK) on success
CVC_ERROR(CVC_E_WRONGOBJECT) if Dncf is not a CVDNCFINDER object
CVC_ERROR(CVC_E_PARAMETER) if some values in params are out of range
CVC_ERROR(CVC_E_ERROR) else

See also: CVDNCGetSearchParams


Contact Legal Privacy Statement	Copyright © 2024.
	All rights reserved.

Data Structures

Typedefs

Functions

Detailed Description

Typedef Documentation

◆ CVDNCFINDER

◆ CVDNCRESULTS

◆ CVDNCTeachCallback

Function Documentation

◆ CVDNCCreateFromFile()

◆ CVDNCCreateFromFileW()

◆ CVDNCCreateFromSamples()

◆ CVDNCFind()

◆ CVDNCGetNumResults()

◆ CVDNCGetResult()

◆ CVDNCGetSearchParams()

◆ CVDNCGetTeachParams()

◆ CVDNCSaveClassifier()

◆ CVDNCSaveClassifierW()

◆ CVDNCSetSearchParams()