Class representing 1 to 4-channel matrix of float or uchar, stored on CPU and/or GPU side. More...

Functions
	Mat ()
	Default constructor. More...

	Mat (size_t width, size_t height, MAT_TYPE mat_type, MEM memory_type=MEM::CPU)
	Constructor. More...

	Mat (size_t width, size_t height, MAT_TYPE mat_type, sl::uchar1 *ptr, size_t step, MEM memory_type=MEM::CPU)
	Constructor from an existing data pointer. More...

	Mat (size_t width, size_t height, MAT_TYPE mat_type, sl::uchar1 ptr_cpu, size_t step_cpu, sl::uchar1 ptr_gpu, size_t step_gpu)
	Constructor from two existing data pointers, CPU and GPU. More...

	Mat (Resolution resolution, MAT_TYPE mat_type, MEM memory_type=MEM::CPU)
	Constructor from an sl::Resolution. More...

	Mat (Resolution resolution, MAT_TYPE mat_type, sl::uchar1 *ptr, size_t step, MEM memory_type=MEM::CPU)
	Constructor from an existing data pointer. More...

	Mat (Resolution resolution, MAT_TYPE mat_type, sl::uchar1 ptr_cpu, size_t step_cpu, sl::uchar1 ptr_gpu, size_t step_gpu)
	Constructor from two existing data pointers, CPU and GPU. More...

	Mat (const Mat &mat)
	Copy constructor (shallow copy). More...

void	alloc (size_t width, size_t height, MAT_TYPE mat_type, MEM memory_type=MEM::CPU)
	Allocates the sl::Mat memory. More...

void	alloc (Resolution resolution, MAT_TYPE mat_type, MEM memory_type=MEM::CPU)
	Allocates the sl::Mat memory. More...

	~Mat ()
	Destructor. More...

void	free (MEM memory_type=MEM::CPU\|MEM::GPU)
	Free the owned memory. More...

Mat &	operator= (const Mat &that)
	Performs a shallow copy. More...

ERROR_CODE	updateCPUfromGPU (cudaStream_t stream=0)
	Downloads data from DEVICE (GPU) to HOST (CPU), if possible. More...

ERROR_CODE	updateGPUfromCPU (cudaStream_t stream=0, int GPU_id=0)
	Uploads data from HOST (CPU) to DEVICE (GPU), if possible. More...

ERROR_CODE	copyTo (Mat &dst, COPY_TYPE cpyType=COPY_TYPE::CPU_CPU, cudaStream_t stream=0) const
	Copies data an other sl::Mat (deep copy). More...

ERROR_CODE	setFrom (const Mat &src, COPY_TYPE cpyType=COPY_TYPE::CPU_CPU, cudaStream_t stream=0)
	Copies data from an other sl::Mat (deep copy). More...

ERROR_CODE	read (const String &filePath)
	Reads an image from a file. More...

ERROR_CODE	write (const String &filePath, sl::MEM memory_type=sl::MEM::CPU, int compression_level=-1)
	Writes the sl::Mat (only if sl::MEM::CPU is available) into a file defined by its extension. More...

template<typename T >
ERROR_CODE	setTo (const T &value, const sl::MEM memory_type=MEM::CPU, cudaStream_t stream=0)
	Fills the sl::Mat with the given value. More...

template<typename N >
ERROR_CODE	setValue (const size_t x, const size_t y, const N &value, const sl::MEM memory_type=MEM::CPU)
	Sets a value to a specific point in the matrix. More...

template<typename N >
ERROR_CODE	getValue (const size_t x, const size_t y, N *value, const MEM memory_type=MEM::CPU) const
	Returns the value of a specific point in the matrix. More...

size_t	getWidth () const
	Returns the width of the matrix. More...

size_t	getHeight () const
	Returns the height of the matrix. More...

Resolution	getResolution () const
	Returns the resolution (width and height) of the matrix. More...

size_t	getChannels () const
	Returns the number of values stored in one pixel. More...

MAT_TYPE	getDataType () const
	Returns the format of the matrix. More...

MEM	getMemoryType () const
	Returns the type of memory (CPU and/or GPU). More...

template<typename N >
N *	getPtr (const MEM memory_type=MEM::CPU) const
	Returns the CPU or GPU data pointer. More...

size_t	getStepBytes (const MEM memory_type=MEM::CPU) const
	Returns the memory step in bytes (size of one pixel row). More...

template<typename N >
size_t	getStep (MEM memory_type=MEM::CPU) const
	Returns the memory step in number of elements (size in one pixel row). More...

size_t	getStep (MEM memory_type=MEM::CPU) const
	Returns the memory step in number of elements (size in one pixel row). More...

size_t	getPixelBytes () const
	Returns the size of one pixel in bytes. More...

size_t	getWidthBytes () const
	Returns the size of a row in bytes. More...

String	getInfos ()
	Returns the information about the sl::Mat into a sl::String. More...

bool	isInit () const
	Returns whether the sl::Mat is initialized or not. More...

bool	isMemoryOwner () const
	Returns whether the sl::Mat is the owner of the memory it accesses. More...

ERROR_CODE	clone (const Mat &src)
	Duplicates a sl::Mat by copy (deep copy). More...

ERROR_CODE	move (Mat &dst)
	Moves the data of the sl::Mat to another sl::Mat. More...

Static Functions
static void	swap (Mat &mat1, Mat &mat2)
	Swaps the content of the provided sl::Mat (only swaps the pointers, no data copy). More...

Attributes
String	name = "n/a"
	Variable used in verbose mode to indicate which sl::Mat is printing informations. More...

bool	verbose = false
	Whether the sl::Mat can display informations. More...

Timestamp	timestamp = 0
	Timestamp of the last manipulation of the data of the matrix by a method/function. More...

Detailed Description

Class representing 1 to 4-channel matrix of float or uchar, stored on CPU and/or GPU side.

This class is defined in a row-major order, meaning that for an image buffer, the rows are stored consecutively from top to bottom.

The CPU and GPU buffer aren't automatically synchronized for performance reasons, you can use updateCPUfromGPU / updateGPUfromCPU to do it.
If you are using the GPU side of the sl::Mat object, you need to make sure to call free before destroying the sl::Camera object.
The destruction of the sl::Camera object deletes the CUDA context needed to free the GPU sl::Mat memory.

Constructor and Destructor

◆ Mat() [1/8]

Mat ( )

Default constructor.

The sl::Mat is empty (width = height = 0).

◆ Mat() [2/8]

Mat	(	size_t	width,
		size_t	height,
		MAT_TYPE	mat_type,
		MEM	memory_type = `MEM::CPU`
	)

Constructor.

Initializes the sl::Mat and allocates the requested memory by calling alloc().

Parameters

width	: Width of the matrix in pixels.
height	: Height of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

◆ Mat() [3/8]

Mat	(	size_t	width,
		size_t	height,
		MAT_TYPE	mat_type,
		sl::uchar1 *	ptr,
		size_t	step,
		MEM	memory_type = `MEM::CPU`
	)

Constructor from an existing data pointer.

Initializes the sl::Mat but does not allocate the memory.

Parameters

width	: Width of the matrix in pixels.
height	: Height of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
ptr	: Pointer to the data array (CPU or GPU).
step	: Step of the data array (bytes size of one pixel row).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

◆ Mat() [4/8]

Mat	(	size_t	width,
		size_t	height,
		MAT_TYPE	mat_type,
		sl::uchar1 *	ptr_cpu,
		size_t	step_cpu,
		sl::uchar1 *	ptr_gpu,
		size_t	step_gpu
	)

Constructor from two existing data pointers, CPU and GPU.

Initializes the sl::Mat but does not allocate the memory.

Parameters

width	: Width of the matrix in pixels.
height	: Height of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
ptr_cpu	: CPU pointer to the data array.
step_cpu	: Step of the CPU data array (bytes size of one pixel row).
ptr_gpu	: GPU pointer to the data array.
step_gpu	: Step of the GPU data array (bytes size of one pixel row).

◆ Mat() [5/8]

Mat	(	Resolution	resolution,
		MAT_TYPE	mat_type,
		MEM	memory_type = `MEM::CPU`
	)

Constructor from an sl::Resolution.

Initializes the sl::Mat and allocates the requested memory by calling alloc().

Parameters

resolution	: Size of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

◆ Mat() [6/8]

Mat	(	Resolution	resolution,
		MAT_TYPE	mat_type,
		sl::uchar1 *	ptr,
		size_t	step,
		MEM	memory_type = `MEM::CPU`
	)

Constructor from an existing data pointer.

Initializes the sl::Mat but does not allocate the memory.

Parameters

resolution	: Size of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
ptr	: Pointer to the data array (CPU or GPU).
step	: Step of the data array (bytes size of one pixel row).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

◆ Mat() [7/8]

Mat	(	Resolution	resolution,
		MAT_TYPE	mat_type,
		sl::uchar1 *	ptr_cpu,
		size_t	step_cpu,
		sl::uchar1 *	ptr_gpu,
		size_t	step_gpu
	)

Constructor from two existing data pointers, CPU and GPU.

Initializes the sl::Mat but does not allocate the memory.

Parameters

resolution	: Size of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
ptr_cpu	: CPU pointer to the data array.
step_cpu	: Step of the CPU data array (bytes size of one pixel row).
ptr_gpu	: GPU pointer to the data array.
step_gpu	: Step of the GPU data array (bytes size of one pixel row).

◆ Mat() [8/8]

Mat ( const Mat & mat )

Copy constructor (shallow copy).

Initializes the sl::Mat but does not allocate the memory.

Parameters

mat	: Reference to the sl::Mat to copy.

◆ ~Mat()

~Mat ( )

Destructor.

This method calls free to release owned memory.

Functions

◆ alloc() [1/2]

void alloc	(	size_t	width,
		size_t	height,
		MAT_TYPE	mat_type,
		MEM	memory_type = `MEM::CPU`
	)

Allocates the sl::Mat memory.

Parameters

width	: Width of the matrix in pixels.
height	: Height of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

Warning: It erases previously allocated memory.

◆ alloc() [2/2]

void alloc	(	Resolution	resolution,
		MAT_TYPE	mat_type,
		MEM	memory_type = `MEM::CPU`
	)

Allocates the sl::Mat memory.

Parameters

resolution	: Size of the matrix in pixels.
mat_type	: Type of the matrix (sl::MAT_TYPE::F32_C1, sl::MAT_TYPE::U8_C4, etc.).
memory_type	: Where the buffer will be stored (sl::MEM::CPU and/or sl::MEM::GPU).

Warning: It erases previously allocated memory.

◆ free()

void free ( MEM memory_type = MEM::CPU|MEM::GPU )

Free the owned memory.

Parameters

memory_type : Specifies whether you want to free the sl::MEM::CPU and/or sl::MEM::GPU memory.

◆ operator=()

Mat& operator= ( const Mat & that )

Performs a shallow copy.

This method does not copy the data array, it only copies the pointer.

Parameters

that	: sl::Mat to be copied.

Returns: The new sl::Mat object which point to the same data as that.

◆ updateCPUfromGPU()

ERROR_CODE updateCPUfromGPU ( cudaStream_t stream = 0 )

Downloads data from DEVICE (GPU) to HOST (CPU), if possible.

Parameters

stream : Specifies the GPU stream to be used to enable asynchronous overlapping. Default: 0 (synchronous)

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: If no CPU or GPU memory are available for this sl::Mat, some are directly allocated.; If verbose is set to true, you have information in case of failure.

◆ updateGPUfromCPU()

ERROR_CODE updateGPUfromCPU	(	cudaStream_t	stream = `0`,
		int	GPU_id = `0`
	)

Uploads data from HOST (CPU) to DEVICE (GPU), if possible.

Parameters

stream : Specifies the GPU stream to be used to enable asynchronous overlapping. Default: 0 (synchronous)

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: If no CPU or GPU memory are available for this sl::Mat, some are directly allocated.; If verbose is set to true, you have information in case of failure.

◆ copyTo()

ERROR_CODE copyTo	(	Mat &	dst,
		COPY_TYPE	cpyType = `COPY_TYPE::CPU_CPU`,
		cudaStream_t	stream = `0`
	)		const

Copies data an other sl::Mat (deep copy).

Parameters

dst	: sl::Mat where the data will be copied.
cpyType	: Specifies the memory that will be used for the copy.
stream	: Specifies the GPU stream to be used to enable asynchronous overlapping. Default: 0 (synchronous)

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: If the destination is not allocated or or doesn't have a compatible sl::MAT_TYPE or sl::Resolution, current memory is freed and new memory is directly allocated.

◆ setFrom()

ERROR_CODE setFrom	(	const Mat &	src,
		COPY_TYPE	cpyType = `COPY_TYPE::CPU_CPU`,
		cudaStream_t	stream = `0`
	)

Copies data from an other sl::Mat (deep copy).

Parameters

src	: sl::Mat where the data will be copied from.
cpyType	: Specifies the memory that will be used for the copy.
stream	: Specifies the GPU stream to be used to enable asynchronous overlapping. Default: 0 (synchronous)

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: If the destination is not allocated or or doesn't have a compatible sl::MAT_TYPE or sl::Resolution, current memory is freed and new memory is directly allocated.

◆ read()

ERROR_CODE read ( const String & filePath )

Reads an image from a file.

Parameters

filePath : Path of the file to read (including the name and extension).

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: This method only support images such as JPG or PNG, and can't load float format such as PCD, PLY, etc.

◆ write()

ERROR_CODE write	(	const String &	filePath,
		sl::MEM	memory_type = `sl::MEM::CPU`,
		int	compression_level = `-1`
	)

Writes the sl::Mat (only if sl::MEM::CPU is available) into a file defined by its extension.

Parameters

filePath	: Path of the file to write in (including the name and extension).
memory_type	: Memory type (CPU or GPU) of the sl::Mat.
compression_level	: Level of compression between 0 (lowest compression == highest size == highest quality(jpg)) and 100 (highest compression == lowest size == lowest quality(jpg)).

Note: Specific/default value for compression_level = -1 : This will set the default quality for PNG(30) or JPEG(5).; compression_level is only supported for U8_Cx sl::MAT_TYPE.

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note

Supported sl::MAT_TYPE are:

sl::MAT_TYPE::F32_C1 for PNG/PFM/PGM
sl::MAT_TYPE::F32_C3 for PCD/PLY/VTK/XYZ
sl::MAT_TYPE::F32_C4 for PCD/PLY/VTK/WYZ
sl::MAT_TYPE::U8_C1 for PNG/JPG
sl::MAT_TYPE::U8_C3 for PNG/JPG
sl::MAT_TYPE::U8_C4 for PNG/JPG

◆ setTo()

ERROR_CODE setTo	(	const T &	value,
		const sl::MEM	memory_type = `MEM::CPU`,
		cudaStream_t	stream = `0`
	)

Fills the sl::Mat with the given value.

This method overwrites all the matrix.

Parameters

value	: Value to be copied all over the matrix.
memory_type	: Which buffer to fill, CPU and/or GPU.
stream	: a cuda stream to put the compute to (default 0).

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: This method is templated for sl::uchar1, sl::uchar2, sl::uchar3, sl::uchar4, sl::float1, sl::float2, sl::float3, sl::float4, sl::char4, sl::ushort1.

◆ setValue()

ERROR_CODE setValue	(	const size_t	x,
		const size_t	y,
		const N &	value,
		const sl::MEM	memory_type = `MEM::CPU`
	)

Sets a value to a specific point in the matrix.

Parameters

x	: Column of the point to change.
y	: Row of the point to change.
value	: Value to be set.
memory_type	: Which memory will be updated.

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: This method is templated for sl::uchar1, sl::uchar2, sl::uchar3, sl::uchar4, sl::float1, sl::float2, sl::float3, sl::float4, sl::char4, sl::ushort1.

Warning: Not efficient for sl::MEM::GPU, use it on sparse data.

◆ getValue()

ERROR_CODE getValue	(	const size_t	x,
		const size_t	y,
		N *	value,
		const MEM	memory_type = `MEM::CPU`
	)		const

Returns the value of a specific point in the matrix.

Parameters

x	: Column of the point to get the value from.
y	: Row of the point to get the value from.
memory_type	: Which memory should be read.

Returns: sl::ERROR_CODE::SUCCESS if everything went well, sl::ERROR_CODE::FAILURE otherwise.

Note: This method is templated for sl::uchar1, sl::uchar2, sl::uchar3, sl::uchar4, sl::float1, sl::float2, sl::float3, sl::float4, sl::char4, sl::ushort1.

Warning: Not efficient for sl::MEM::GPU, use it on sparse data.

◆ getWidth()

size_t getWidth ( ) const

inline

Returns the width of the matrix.

Returns: Width of the matrix in pixels.

◆ getHeight()

size_t getHeight ( ) const

inline

Returns the height of the matrix.

Returns: Height of the matrix in pixels.

◆ getResolution()

Resolution getResolution ( ) const

inline

Returns the resolution (width and height) of the matrix.

Returns: Resolution of the matrix in pixels.

◆ getChannels()

size_t getChannels ( ) const

inline

Returns the number of values stored in one pixel.

Returns: Number of values in a pixel.

◆ getDataType()

MAT_TYPE getDataType ( ) const

inline

Returns the format of the matrix.

Returns: Format of the current sl::Mat.

◆ getMemoryType()

MEM getMemoryType ( ) const

inline

Returns the type of memory (CPU and/or GPU).

Returns: Type of allocated memory.

◆ getPtr()

N* getPtr ( const MEM memory_type = MEM::CPU ) const

Returns the CPU or GPU data pointer.

Parameters

memory_type : Specifies whether you want sl::MEM::CPU or sl::MEM::GPU.

Returns: The pointer of the Mat data.

◆ getStepBytes()

size_t getStepBytes ( const MEM memory_type = MEM::CPU ) const

Returns the memory step in bytes (size of one pixel row).

Parameters

memory_type : Specifies whether you want sl::MEM::CPU or sl::MEM::GPU step.

Returns: The step in bytes of the specified memory.

◆ getStep() [1/2]

size_t getStep ( MEM memory_type = MEM::CPU ) const

inline

Returns the memory step in number of elements (size in one pixel row).

Parameters

memory_type : Specifies whether you want sl::MEM::CPU or sl::MEM::GPU step.

Returns: The step in number of elements.

◆ getStep() [2/2]

size_t getStep ( MEM memory_type = MEM::CPU ) const

inline

Returns the memory step in number of elements (size in one pixel row).

Parameters

memory_type : Specifies whether you want sl::MEM::CPU or sl::MEM::GPU step.

Returns: The step in number of elements.

◆ getPixelBytes()

size_t getPixelBytes ( ) const

inline

Returns the size of one pixel in bytes.

Returns: Size of a pixel in bytes.

◆ getWidthBytes()

size_t getWidthBytes ( ) const

inline

Returns the size of a row in bytes.

Returns: Size of a row in bytes.

◆ getInfos()

String getInfos ( )

Returns the information about the sl::Mat into a sl::String.

Returns: String containing the sl::Mat information.

◆ isInit()

bool isInit ( ) const

inline

Returns whether the sl::Mat is initialized or not.

Returns: True if current sl::Mat has been allocated (by the constructor or therefore).

◆ isMemoryOwner()

bool isMemoryOwner ( ) const

inline

Returns whether the sl::Mat is the owner of the memory it accesses.

If not, the memory won't be freed if the sl::Mat is destroyed.

Returns: True if the sl::Mat is owning its memory, else false.

◆ clone()

ERROR_CODE clone ( const Mat & src )

Duplicates a sl::Mat by copy (deep copy).

Parameters

src	: Reference of the sl::Mat to copy.

This method copies the data array(s) and it marks the new sl::Mat as the memory owner.

◆ move()

ERROR_CODE move ( Mat & dst )

Moves the data of the sl::Mat to another sl::Mat.

This method gives the attribute of the current s::Mat to the specified one. (No copy.)

Parameters

dst	: Reference to the sl::Mat to move to.

Note: : The current sl::Mat is then no more usable since its loose its attributes.

◆ swap()

static void swap	(	Mat &	mat1,
		Mat &	mat2
	)

static

Swaps the content of the provided sl::Mat (only swaps the pointers, no data copy).

This method swaps the pointers of the given sl::Mat.

Parameters

mat1	: First matrix to swap.
mat2	: Second matrix to swap.

Variables

◆ name

String name = "n/a"

Variable used in verbose mode to indicate which sl::Mat is printing informations.

Default set to "n/a" to avoid empty string if not filled.

◆ verbose

bool verbose = false

Whether the sl::Mat can display informations.

◆ timestamp

Timestamp timestamp = 0

Timestamp of the last manipulation of the data of the matrix by a method/function.

Functions

Static Functions

Attributes

Detailed Description

Constructor and Destructor

◆ Mat() [1/8]

◆ Mat() [2/8]

◆ Mat() [3/8]

◆ Mat() [4/8]

◆ Mat() [5/8]

◆ Mat() [6/8]

◆ Mat() [7/8]

◆ Mat() [8/8]

◆ ~Mat()

Functions

◆ alloc() [1/2]

◆ alloc() [2/2]

◆ free()

◆ operator=()

◆ updateCPUfromGPU()

◆ updateGPUfromCPU()

◆ copyTo()

◆ setFrom()

◆ read()

◆ write()

◆ setTo()

◆ setValue()

◆ getValue()

◆ getWidth()

◆ getHeight()

◆ getResolution()

◆ getChannels()

◆ getDataType()

◆ getMemoryType()

◆ getPtr()

◆ getStepBytes()

◆ getStep() [1/2]

◆ getStep() [2/2]

◆ getPixelBytes()

◆ getWidthBytes()

◆ getInfos()

◆ isInit()

◆ isMemoryOwner()

◆ clone()

◆ move()

◆ swap()

Variables

◆ name

◆ verbose

◆ timestamp