Base storage class for GPU memory with reference counting. More...

#include <cuda.hpp>

Public Member Functions
	GpuMat (Allocator *allocator=defaultAllocator())
	default constructor
	GpuMat (int rows, int cols, int type, Allocator *allocator=defaultAllocator())
	constructs GpuMat of the specified size and type
	GpuMat (int rows, int cols, int type, Scalar s, Allocator *allocator=defaultAllocator())
	constucts GpuMat and fills it with the specified value _s
	GpuMat (const GpuMat &m)
	copy constructor
	GpuMat (int rows, int cols, int type, void *data, size_t step=Mat::AUTO_STEP)
	constructor for GpuMat headers pointing to user-allocated data
	GpuMat (const GpuMat &m, Range rowRange, Range colRange)
	creates a GpuMat header for a part of the bigger matrix
	GpuMat (InputArray arr, Allocator *allocator=defaultAllocator())
	builds GpuMat from host memory (Blocking call)
	~GpuMat ()
	destructor - calls release()
GpuMat &	operator= (const GpuMat &m)
	assignment operators
void	create (int rows, int cols, int type)
	allocates new GpuMat data unless the GpuMat already has specified size and type
void	release ()
	decreases reference counter, deallocate the data when reference counter reaches 0
void	swap (GpuMat &mat)
	swaps with other smart pointer
void	upload (InputArray arr)
	pefroms upload data to GpuMat (Blocking call)
void	upload (InputArray arr, Stream &stream)
	pefroms upload data to GpuMat (Non-Blocking call)
void	download (OutputArray dst) const
	pefroms download data from device to host memory (Blocking call)
void	download (OutputArray dst, Stream &stream) const
	pefroms download data from device to host memory (Non-Blocking call)
GpuMat	clone () const
	returns deep copy of the GpuMat, i.e. the data is copied
void	copyTo (OutputArray dst) const
	copies the GpuMat content to device memory (Blocking call)
void	copyTo (OutputArray dst, Stream &stream) const
	copies the GpuMat content to device memory (Non-Blocking call)
void	copyTo (OutputArray dst, InputArray mask) const
	copies those GpuMat elements to "m" that are marked with non-zero mask elements (Blocking call)
void	copyTo (OutputArray dst, InputArray mask, Stream &stream) const
	copies those GpuMat elements to "m" that are marked with non-zero mask elements (Non-Blocking call)
GpuMat &	setTo (Scalar s)
	sets some of the GpuMat elements to s (Blocking call)
GpuMat &	setTo (Scalar s, Stream &stream)
	sets some of the GpuMat elements to s (Non-Blocking call)
GpuMat &	setTo (Scalar s, InputArray mask)
	sets some of the GpuMat elements to s, according to the mask (Blocking call)
GpuMat &	setTo (Scalar s, InputArray mask, Stream &stream)
	sets some of the GpuMat elements to s, according to the mask (Non-Blocking call)
void	convertTo (OutputArray dst, int rtype) const
	converts GpuMat to another datatype (Blocking call)
void	convertTo (OutputArray dst, int rtype, Stream &stream) const
	converts GpuMat to another datatype (Non-Blocking call)
void	convertTo (OutputArray dst, int rtype, double alpha, double beta=0.0) const
	converts GpuMat to another datatype with scaling (Blocking call)
void	convertTo (OutputArray dst, int rtype, double alpha, Stream &stream) const
	converts GpuMat to another datatype with scaling (Non-Blocking call)
void	convertTo (OutputArray dst, int rtype, double alpha, double beta, Stream &stream) const
	converts GpuMat to another datatype with scaling (Non-Blocking call)
uchar *	ptr (int y=0)
	returns pointer to y-th row
template<typename _Tp >
_Tp *	ptr (int y=0)
	template version of the above method
GpuMat	row (int y) const
	returns a new GpuMat header for the specified row
GpuMat	col (int x) const
	returns a new GpuMat header for the specified column
GpuMat	rowRange (int startrow, int endrow) const
	... for the specified row span
GpuMat	colRange (int startcol, int endcol) const
	... for the specified column span
GpuMat	operator() (Range rowRange, Range colRange) const
	extracts a rectangular sub-GpuMat (this is a generalized form of row, rowRange etc.)
GpuMat	reshape (int cn, int rows=0) const
	creates alternative GpuMat header for the same data, with different number of channels and/or different number of rows
void	locateROI (Size &wholeSize, Point &ofs) const
	locates GpuMat header within a parent GpuMat
GpuMat &	adjustROI (int dtop, int dbottom, int dleft, int dright)
	moves/resizes the current GpuMat ROI inside the parent GpuMat
bool	isContinuous () const
	returns true iff the GpuMat data is continuous (i.e.
size_t	elemSize () const
	returns element size in bytes
size_t	elemSize1 () const
	returns the size of element channel in bytes
int	type () const
	returns element type
int	depth () const
	returns element type
int	channels () const
	returns number of channels
size_t	step1 () const
	returns step/elemSize1()
Size	size () const
	returns GpuMat size : width == number of columns, height == number of rows
bool	empty () const
	returns true if GpuMat data is NULL
Static Public Member Functions
static Allocator *	defaultAllocator ()
	default allocator
Data Fields
int	flags
int	rows
	the number of rows and columns
size_t	step
	a distance between successive rows in bytes; includes the gap if any
uchar *	data
	pointer to the data
int *	refcount
	pointer to the reference counter; when GpuMat points to user-allocated data, the pointer is NULL
uchar *	datastart
	helper fields used in locateROI and adjustROI
Allocator *	allocator
	allocator

Detailed Description

Base storage class for GPU memory with reference counting.

Its interface matches the Mat interface with the following limitations:

no arbitrary dimensions support (only 2D)
no functions that return references to their data (because references on GPU are not valid for CPU)
no expression templates technique support

Beware that the latter limitation may lead to overloaded matrix operators that cause memory allocations. The GpuMat class is convertible to cuda::PtrStepSz and cuda::PtrStep so it can be passed directly to the kernel.

Note:: In contrast with Mat, in most cases GpuMat::isContinuous() == false . This means that rows are aligned to a size depending on the hardware. Single-row GpuMat is always a continuous matrix.; You are not recommended to leave static or global GpuMat variables allocated, that is, to rely on its destructor. The destruction order of such variables and CUDA context is undefined. GPU memory release function returns error if the CUDA context has been destroyed before.

See also:: Mat

Definition at line 96 of file cuda.hpp.

Constructor & Destructor Documentation

GpuMat ( Allocator * allocator = defaultAllocator() ) [explicit]

default constructor

GpuMat	(	int	rows,
		int	cols,
		int	type,
		Allocator *	allocator = `defaultAllocator()`
	)

constructs GpuMat of the specified size and type

GpuMat	(	int	rows,
		int	cols,
		int	type,
		Scalar	s,
		Allocator *	allocator = `defaultAllocator()`
	)

constucts GpuMat and fills it with the specified value _s

GpuMat ( const GpuMat & m )

copy constructor

GpuMat	(	int	rows,
		int	cols,
		int	type,
		void *	data,
		size_t	step = `Mat::AUTO_STEP`
	)

constructor for GpuMat headers pointing to user-allocated data

Definition at line 49 of file cuda_gpu_mat.cpp.

GpuMat	(	const GpuMat &	m,
		Range	rowRange,
		Range	colRange
	)

creates a GpuMat header for a part of the bigger matrix

Definition at line 101 of file cuda_gpu_mat.cpp.

GpuMat	(	InputArray	arr,
		Allocator *	allocator = `defaultAllocator()`
	)		`[explicit]`

builds GpuMat from host memory (Blocking call)

~GpuMat ( )

destructor - calls release()

Member Function Documentation

GpuMat & adjustROI	(	int	dtop,
		int	dbottom,
		int	dleft,
		int	dright
	)

moves/resizes the current GpuMat ROI inside the parent GpuMat

Definition at line 230 of file cuda_gpu_mat.cpp.

int channels ( ) const

returns number of channels

GpuMat clone ( ) const

returns deep copy of the GpuMat, i.e. the data is copied

GpuMat col ( int x ) const

returns a new GpuMat header for the specified column

GpuMat colRange	(	int	startcol,
		int	endcol
	)		const

... for the specified column span

void convertTo	(	OutputArray	dst,
		int	rtype
	)		const

converts GpuMat to another datatype (Blocking call)

void convertTo	(	OutputArray	dst,
		int	rtype,
		Stream &	stream
	)		const

converts GpuMat to another datatype (Non-Blocking call)

Definition at line 503 of file cuda_gpu_mat.cpp.

void convertTo	(	OutputArray	dst,
		int	rtype,
		double	alpha,
		double	beta = `0.0`
	)		const

converts GpuMat to another datatype with scaling (Blocking call)

void convertTo	(	OutputArray	dst,
		int	rtype,
		double	alpha,
		Stream &	stream
	)		const

converts GpuMat to another datatype with scaling (Non-Blocking call)

void convertTo	(	OutputArray	dst,
		int	rtype,
		double	alpha,
		double	beta,
		Stream &	stream
	)		const

converts GpuMat to another datatype with scaling (Non-Blocking call)

Definition at line 511 of file cuda_gpu_mat.cpp.

void copyTo	(	OutputArray	dst,
		InputArray	mask,
		Stream &	stream
	)		const

copies those GpuMat elements to "m" that are marked with non-zero mask elements (Non-Blocking call)

Definition at line 478 of file cuda_gpu_mat.cpp.

void copyTo ( OutputArray dst ) const

copies the GpuMat content to device memory (Blocking call)

Definition at line 465 of file cuda_gpu_mat.cpp.

void copyTo	(	OutputArray	dst,
		Stream &	stream
	)		const

copies the GpuMat content to device memory (Non-Blocking call)

Definition at line 471 of file cuda_gpu_mat.cpp.

void copyTo	(	OutputArray	dst,
		InputArray	mask
	)		const

copies those GpuMat elements to "m" that are marked with non-zero mask elements (Blocking call)

void create	(	int	rows,
		int	cols,
		int	type
	)

allocates new GpuMat data unless the GpuMat already has specified size and type

Definition at line 427 of file cuda_gpu_mat.cpp.

GpuMat::Allocator * defaultAllocator ( ) [static]

default allocator

Definition at line 416 of file cuda_gpu_mat.cpp.

int depth ( ) const

returns element type

void download ( OutputArray dst ) const

pefroms download data from device to host memory (Blocking call)

Definition at line 452 of file cuda_gpu_mat.cpp.

void download	(	OutputArray	dst,
		Stream &	stream
	)		const

pefroms download data from device to host memory (Non-Blocking call)

Definition at line 458 of file cuda_gpu_mat.cpp.

size_t elemSize ( ) const

returns element size in bytes

size_t elemSize1 ( ) const

returns the size of element channel in bytes

bool empty ( ) const

returns true if GpuMat data is NULL

bool isContinuous ( ) const

returns true iff the GpuMat data is continuous (i.e.

when there are no gaps between successive rows)

void locateROI	(	Size &	wholeSize,
		Point &	ofs
	)		const

locates GpuMat header within a parent GpuMat

Definition at line 204 of file cuda_gpu_mat.cpp.

GpuMat operator()	(	Range	rowRange,
		Range	colRange
	)		const

extracts a rectangular sub-GpuMat (this is a generalized form of row, rowRange etc.)

GpuMat& operator= ( const GpuMat & m )

assignment operators

_Tp* ptr ( int y = 0 )

template version of the above method

uchar* ptr ( int y = 0 )

returns pointer to y-th row

void release ( void )

decreases reference counter, deallocate the data when reference counter reaches 0

Definition at line 435 of file cuda_gpu_mat.cpp.

GpuMat reshape	(	int	cn,
		int	rows = `0`
	)		const

creates alternative GpuMat header for the same data, with different number of channels and/or different number of rows

Definition at line 161 of file cuda_gpu_mat.cpp.

GpuMat row ( int y ) const

returns a new GpuMat header for the specified row

GpuMat rowRange	(	int	startrow,
		int	endrow
	)		const

... for the specified row span

GpuMat & setTo	(	Scalar	s,
		InputArray	mask,
		Stream &	stream
	)

sets some of the GpuMat elements to s, according to the mask (Non-Blocking call)

Definition at line 494 of file cuda_gpu_mat.cpp.

GpuMat& setTo	(	Scalar	s,
		InputArray	mask
	)

sets some of the GpuMat elements to s, according to the mask (Blocking call)

GpuMat & setTo	(	Scalar	s,
		Stream &	stream
	)

sets some of the GpuMat elements to s (Non-Blocking call)

Definition at line 486 of file cuda_gpu_mat.cpp.

GpuMat& setTo ( Scalar s )

sets some of the GpuMat elements to s (Blocking call)

Size size ( ) const

returns GpuMat size : width == number of columns, height == number of rows

size_t step1 ( ) const

returns step/elemSize1()

void swap ( GpuMat & mat )

swaps with other smart pointer

int type ( ) const

returns element type

void upload ( InputArray arr )

pefroms upload data to GpuMat (Blocking call)

Definition at line 439 of file cuda_gpu_mat.cpp.

void upload	(	InputArray	arr,
		Stream &	stream
	)

pefroms upload data to GpuMat (Non-Blocking call)

Definition at line 445 of file cuda_gpu_mat.cpp.

Field Documentation

Allocator* allocator

allocator

Definition at line 303 of file cuda.hpp.

uchar* data

pointer to the data

Definition at line 292 of file cuda.hpp.

uchar* datastart

helper fields used in locateROI and adjustROI

Definition at line 299 of file cuda.hpp.

int flags

includes several bit-fields:

the magic signature
continuity flag
depth
number of channels

Definition at line 283 of file cuda.hpp.

int* refcount

pointer to the reference counter; when GpuMat points to user-allocated data, the pointer is NULL

Definition at line 296 of file cuda.hpp.

int rows

the number of rows and columns

Definition at line 286 of file cuda.hpp.

size_t step

a distance between successive rows in bytes; includes the gap if any

Definition at line 289 of file cuda.hpp.

Type:	Program
Created:	26 Jul 2017
Imports:	3
Forks:	0
Commits:	168
Dependents:	0
Dependencies:	0
Followers:	9

Important changes to repositories hosted on mbed.com

GpuMat Class Reference
[Data Structures]

Public Member Functions

Static Public Member Functions

Data Fields

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Field Documentation

Repository toolbox

Repository details

Important Information for this Arm website

Important changes to repositories hosted on mbed.com

GpuMat Class Reference [Data Structures]

Public Member Functions

Static Public Member Functions

Data Fields

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Field Documentation

Repository toolbox

Repository details

Important Information for this Arm website

Access Warning

GpuMat Class Reference
[Data Structures]