Frame Class Reference

Frame format and metadata. More...

#include <libfreenect2/frame_listener.hpp>

Public Types
enum	Type { Color = 1 , Ir = 2 , Depth = 4 }
	Available types of frames. More...
enum	Format {   Invalid = 0 , Raw = 1 , Float = 2 , BGRX = 4 ,   RGBX = 5 , Gray = 6 }
	Pixel format. More...

Public Member Functions
	Frame (size_t width_arg, size_t height_arg, size_t bytes_per_pixel_arg, unsigned char *data_arg=NULL)
	Construct a new frame.
	Frame (size_t width_arg, size_t height_arg, size_t bytes_per_pixel_arg, unsigned char *data_arg, Format format_arg)
	Construct a frame with an explicit pixel format.
virtual	~Frame ()

Public Attributes
size_t	width
	Length of a line (in pixels).
size_t	height
	Number of lines in the frame.
size_t	bytes_per_pixel
	Number of bytes in a pixel. If frame format is 'Raw' this is the buffer size.
unsigned char *	data
	Data of the frame (aligned).
uint32_t	timestamp
	Unit: 0.125 millisecond. Usually incrementing by 266 (30Hz) or 533 (15Hz).
uint64_t	arrival_timestamp_us
	Monotonic host time of the first contributing USB transfer, in microseconds.
uint32_t	sequence
	Increasing frame sequence number.
float	exposure
	From 0.5 (very bright) to ~60.0 (fully covered).
float	gain
	From 1.0 (bright) to 1.5 (covered).
float	gamma
	From 1.0 (bright) to 6.4 (covered).
uint32_t	status
	zero if ok; non-zero for errors.
Format	format
	Byte format.

Protected Attributes
unsigned char *	rawdata
	Unaligned start of data.

Detailed Description

Frame format and metadata.

Member Enumeration Documentation

Type

enum Type

Available types of frames.

Enumerator
Color	1920x1080. BGRX or RGBX.
Ir	512x424 float. Range is [0.0, 65535.0].
Depth	512x424 float, unit: millimeter. Non-positive, NaN, and infinity are invalid or missing data.

Format

enum Format

Pixel format.

Enumerator
Invalid	Invalid format.
Raw	Raw bitstream. 'bytes_per_pixel' defines the number of bytes.
Float	A 4-byte float per pixel.
BGRX	4 bytes of B, G, R, and unused per pixel
RGBX	4 bytes of R, G, B, and unused per pixel
Gray	1 byte of gray per pixel

Constructor & Destructor Documentation

Frame() [1/2]

Frame	(	size_t	width_arg,
		size_t	height_arg,
		size_t	bytes_per_pixel_arg,
		unsigned char *	data_arg = NULL )

Construct a new frame.

Parameters

width_arg	Width in pixel
height_arg	Height in pixel
bytes_per_pixel_arg	Bytes per pixel
data_arg	Memory to store frame data. If NULL, new memory is allocated.

Frame() [2/2]

Frame	(	size_t	width_arg,
		size_t	height_arg,
		size_t	bytes_per_pixel_arg,
		unsigned char *	data_arg,
		Format	format_arg )

Construct a frame with an explicit pixel format.

This overload is intended for callers wrapping externally owned image buffers. Consumers such as Registration validate format and will reject the legacy constructor's default Invalid value.

~Frame()

virtual ~Frame ( )

virtual

Member Data Documentation

width

size_t width

Length of a line (in pixels).

height

size_t height

Number of lines in the frame.

bytes_per_pixel

size_t bytes_per_pixel

Number of bytes in a pixel. If frame format is 'Raw' this is the buffer size.

data

unsigned char* data

Data of the frame (aligned).

See also

See Frame::Type for pixel format.

timestamp

uint32_t timestamp

Unit: 0.125 millisecond. Usually incrementing by 266 (30Hz) or 533 (15Hz).

arrival_timestamp_us

uint64_t arrival_timestamp_us

Monotonic host time of the first contributing USB transfer, in microseconds.

sequence

uint32_t sequence

Increasing frame sequence number.

exposure

float exposure

From 0.5 (very bright) to ~60.0 (fully covered).

gain

float gain

From 1.0 (bright) to 1.5 (covered).

gamma

float gamma

From 1.0 (bright) to 6.4 (covered).

status

uint32_t status

zero if ok; non-zero for errors.

format

Format format

Byte format.

Validated by consumers such as Registration::apply() and the vision helpers; the constructor defaults it to Invalid, so callers wrapping their own buffers must set it explicitly.

rawdata

unsigned char* rawdata

protected

Unaligned start of data.