#include <ColorVideo.hh>

Inheritance diagram for ColorVideo:

Collaboration diagram for ColorVideo:

Public Member Functions
	ColorVideo (void)
	ColorVideo (const ColorVideo &cv)
	ColorVideo (int colors, clrmodel colormodel, int rows, int cols, int frames)
virtual	~ColorVideo (void)
virtual void	swapColors (int ch1, int ch2)
void	colormodelVideo (clrmodel model)
Image &	currentFrameChannel (int num=0)
ColorImage &	currentFrame (void)
double	frameRate (void) const
void	frameRate (double fr)
int	current (void) const
void	current (int current)
int	frames (void) const
void	reframe (int frames, bool keephead)
void	resize (int rows, int cols)
void	read (char const *fname, int from=0, int to=0, int rawy=0, int rawx=0, int colors=3, clrmodel cmodel=cm_rgb, int skip=512)
void	read (char const *fname, filetype ftype, int from=0, int to=0, int rawy=0, int rawx=0, int colors=3, clrmodel cmodel=cm_rgb, int skip=512)
void	write (char const *fname, int rawskip=512, bool beautify=false)
void	write (char const *fname, filetype ftype, int rawskip=512, bool beautify=false)
virtual coeff	smax (void) const
virtual coeff	smin (void) const
virtual bool	equals (ColorBuffer &buf) const
virtual bool	epsilons (ColorBuffer &buf, coeff epsilon) const
bool	epsilonsFrames (int f1, int f2, coeff epsilon) const
bool	equalsFrames (int f1, int f2) const
virtual void	truncate (coeff min=0, coeff max=255)
virtual void	beautify (int yoffs=0, int xoffs=0, int ysize=0, int xsize=0)
virtual coeff	averageColor (int fromY=0, int fromX=0, int toY=-1, int toX=-1)
virtual unsigned char *	toPixmap (clrmodel cmodel)
ColorVideo *	clone (void) const
Static Public Member Functions
static void	videoDimensions (const char *fname, int &y, int &x, int &z, int colors, int skip=0)
Protected Member Functions
virtual void	destroy (void)
bool	epsilonsVideo (ColorVideo &cv, coeff epsilon) const
void	init (int colors, clrmodel cmodel, int rows, int cols, int frames)
Protected Attributes
int	m_frames
int	m_current
int	m_xyzsize
VideoArray< coeff > **	m_arrays
ColorImage *	m_colorImage
double	m_frameRate

Detailed Description

An wrapper for videos containing several color components. Objects may import and export different video formats, such as RAW. The color models can be changed (RGB or YUV).

Definition at line 24 of file ColorVideo.hh.

Constructor & Destructor Documentation

ColorVideo::ColorVideo ( void )

Default Constructor. Creates objects and sets actual start values.

ColorVideo::ColorVideo ( const ColorVideo & cv )

Copy Constructor.

Parameters:

cv	the other color video

ColorVideo::ColorVideo	(	int	colors,
		clrmodel	colormodel,
		int	rows,
		int	cols,
		int	frames
	)

Constructor setting a particular size etc. Initialization not done.

Parameters:

colors	the number of colors
colormodel	the initial color model
rows	the number of rows
cols	the number of cols
frames	the number of frames

virtual ColorVideo::~ColorVideo ( void ) [virtual]

Destructor. Frees allocated objects.

Member Function Documentation

virtual coeff ColorVideo::averageColor	(	int	fromY = `0`,
		int	fromX = `0`,
		int	toY = `-1`,
		int	toX = `-1`
	)		`[virtual]`

Return the average color for a rectangular region inside the video drawn from one point within and the second point just outside the region

Parameters:

fromY	the first point's row
fromX	the first point's col
toY	the second point's row (-1 if lower video border)
toX	the second point's col (-1 if right video border)

Returns:: the average color

Reimplemented from ColorBuffer.

virtual void ColorVideo::beautify	(	int	yoffs = `0`,
		int	xoffs = `0`,
		int	ysize = `0`,
		int	xsize = `0`
	)		`[virtual]`

Do an intelligent kind of beautification considering previous transform steps. The original data will be overwritten, so this should only be used on clones or if the video is not going to be transformed afterwise. One region inside the video could be the low-pass subband and should thus be handled separately. If that is desired, the {ysize} and {xsize} parameters should be set to nonzero values. Note that the current implementation is very memory consuming.

Parameters:

ysize	the horizontal size of the region that should be handled separately
xsize	the vertical size of the region that should be handled separately
yoffs	the horizontal offset of the region that should be handled separately
xoffs	the vertical offset of the region that should be handled separately

Implements ColorBuffer.

ColorVideo* ColorVideo::clone ( void ) const

Produce a copy of this video

Returns:: the copy

void ColorVideo::colormodelVideo ( clrmodel model )

Convert the video to a new colormodel. This applies the `ColorBuffer::colormodel()' operation on the whole video instead of only the current frame.

Parameters:

model the new colormodel ({cm_rgb} or {cm_yuv})

int ColorVideo::current ( void ) const [inline]

Returns the number of the current frame.

Returns:: the current frame (counted from 0)

Definition at line 83 of file ColorVideo.hh.

References m_current.

void ColorVideo::current ( int current )

Returns the number of the current frame.

Returns:: the current frame (counted from 0)

ColorImage& ColorVideo::currentFrame ( void ) [inline]

Returns a reference to a color video containing the color channels of the current frame.

Returns:: the frame as a reference to an ColorImage object

Definition at line 71 of file ColorVideo.hh.

References m_colorImage.

Image& ColorVideo::currentFrameChannel ( int num = 0 ) [inline]

Returns a reference to one of the color channels of the current frame. {Don't change the dimensions of single channels unless you know what you're doing!}

Parameters:

num	the number of the color channel

Returns:: the frame as a reference to an VideoFrame object

Definition at line 66 of file ColorVideo.hh.

References ColorBuffer::m_images.

virtual void ColorVideo::destroy ( void ) [protected, virtual]

Deletes allocated objects.

Reimplemented from ColorBuffer.

virtual bool ColorVideo::epsilons	(	ColorBuffer &	buf,
		coeff	epsilon
	)		const `[virtual]`

Rough comparison. See if two videos are similar according to a given {epsilon} (important for floating-point comparisons).

Parameters:

buf	the other {ColorBuffer} object
epsilon	the epsilon

Returns:: if both are identical: {true}, else {false}

Reimplemented from ColorBuffer.

bool ColorVideo::epsilonsFrames	(	int	f1,
		int	f2,
		coeff	epsilon
	)		const

Rough comparison. See if this and another frame are similar according to a given {epsilon} (important for floating-point comparisons).

Exceptions:

invalid_argument one of the frames is out of bounds

Parameters:

f1	the first frame
f2	the second frame
epsilon	the epsilon

Returns:: if both are identical: {true}, else {false}

bool ColorVideo::epsilonsVideo	(	ColorVideo &	cv,
		coeff	epsilon
	)		const `[protected]`

Compares two videos

Parameters:

cv	the other video
epsilon	the min difference

Returns:: true if they equal according to epsilon

virtual bool ColorVideo::equals ( ColorBuffer & buf ) const [virtual]

Compares two videos. Return {true} if both are equal.

Parameters:

buf	The other {ColorBuffer} object

Returns:: if equals: {true}, else {false}

Reimplemented from ColorBuffer.

bool ColorVideo::equalsFrames	(	int	f1,
		int	f2
	)		const

Exact comparison. See if two frames are similar.

Exceptions:

invalid_argument one of the frames is out of bounds

Parameters:

f1	the first frame
f2	the second frame if both are identical: {true}, else {false}

double ColorVideo::frameRate ( void ) const [inline]

Returns the frame rate

Returns:: the frameRate

Definition at line 75 of file ColorVideo.hh.

References m_frameRate.

void ColorVideo::frameRate ( double fr ) [inline]

Sets the frame rate

Parameters:

fr	the frameRate

Definition at line 79 of file ColorVideo.hh.

References m_frameRate.

int ColorVideo::frames ( void ) const [inline]

Returns the number of frames

Returns:: the number of frames

Definition at line 91 of file ColorVideo.hh.

References m_frames.

void ColorVideo::init	(	int	colors,
		clrmodel	cmodel,
		int	rows,
		int	cols,
		int	frames
	)		`[protected]`

Allocate objects and set standard values.

Exceptions:

invalid_argument if any of the arguments does not make sense

Parameters:

colors	the number of colors
cmodel	the initial color model
rows	the number of rows
cols	the number of cols
frames	the number of frames

void ColorVideo::read	(	char const *	fname,
		int	from = `0`,
		int	to = `0`,
		int	rawy = `0`,
		int	rawx = `0`,
		int	colors = `3`,
		clrmodel	cmodel = `cm_rgb`,
		int	skip = `512`
	)

Read the video. All steps independent of the file format will be performed, like testing for file readability etc.

Exceptions:

invalid_argument	invalid file format
ios_base::failure	a read error has occured [not supported by all libraries, so eventually {invalid_argument} instead]

Parameters:

fname	the file name to read from
rawy	the number of rows (only needed for RAW format)
rawx	the number of cols (only needed for RAW format)
colors	the number of colors (if known, necessary for raw videos)
from	the frame to start with (from 0)
to	one greater than the last frame, if equal to `from' then the whole video will be read, if 0, then the video will be read from `from' to the end
cmodel	the color model
skip	the amount of bytes to skip as a header before contents

void ColorVideo::read	(	char const *	fname,
		filetype	ftype,
		int	from = `0`,
		int	to = `0`,
		int	rawy = `0`,
		int	rawx = `0`,
		int	colors = `3`,
		clrmodel	cmodel = `cm_rgb`,
		int	skip = `512`
	)

Read an video. Reads an video from a file using the specified file type. Currently PGM, RAW and PFI are supported.

Exceptions:

invalid_argument the file type is either not supported or could not be determined from the given file name

Parameters:

fname	the file name, if {NULL}, then {stdin}
ftype	the file type
rawy	the number of rows (only needed for RAW format)
rawx	the number of cols (only needed for RAW format)
colors	the number of colors (if known, necessary for raw videos)
from	the frame to start with (from 0)
to	one greater than the last frame, if equal to `from' then the whole video will be read, if 0, then the video will be read from `from' to the end
cmodel	the color model
skip	the amount of bytes to skip as a header before contents

void ColorVideo::reframe	(	int	frames,
		bool	keephead
	)

Sets new number of frames. The old frames will be copied to the new ones as far as they fit in. The new number must be greater than zero.

Exceptions:

invalid_argument the new number of frames is either negative or zero

Parameters:

frames	the new number of frames.
keephead	if shrinking the video, frames will be deleted at the end.

void ColorVideo::resize	(	int	rows,
		int	cols
	)

Resize the video. The old values will be copied to the new dimension as far as they fit in. The new dimensions must all be greater than zero.

Exceptions:

invalid_argument one or both dimensions are either negative or zero

Parameters:

rows	the new number of rows
cols	the new number of cols

virtual coeff ColorVideo::smax ( void ) const [virtual]

Returns the maximum value. Signs will be considered.

Exceptions:

invalid_argument empty video

Returns:: the maximum

Reimplemented from ColorBuffer.

virtual coeff ColorVideo::smin ( void ) const [virtual]

Returns the maximum value. Signs will be considered.

Exceptions:

invalid_argument empty video

Returns:: the maximum

Reimplemented from ColorBuffer.

virtual void ColorVideo::swapColors	(	int	ch1,
		int	ch2
	)		`[virtual]`

Swap two color channels.

Exceptions:

invalid_argument if the channel specs are out of bounds

Parameters:

ch1	the first channel
ch2	the second channel

Reimplemented from ColorBuffer.

virtual unsigned char* ColorVideo::toPixmap ( clrmodel cmodel ) [virtual]

Export the video for other applications. This only works if the video has three channels!

Exceptions:

invalid_argument if the current colormodel or the one passed as parameter is unsupported.

Parameters:

cmodel the color model in which the video is exported. Currently {cm_rgb} and {cm_yuv} are supported.

Returns:: a new buffer containing the video pixels in the format according on the cmodel parameter

Reimplemented from ColorBuffer.

virtual void ColorVideo::truncate	(	coeff	min = `0`,
		coeff	max = `255`
	)		`[virtual]`

Do a dumb kind of beautification by just truncating values to their possible maxima depending on the video format or whatever. The original data will be overwritten, so this should only be used on clones or if the original data is not needed anymore.

Parameters:

min	the minimum allowed value
max	the maximum allowed value

Reimplemented from ColorBuffer.

static void ColorVideo::videoDimensions	(	const char *	fname,
		int &	y,
		int &	x,
		int &	z,
		int	colors,
		int	skip = `0`
	)		`[static]`

Try to determine some of a video file's dimensions: If the file is a raw video, two out of the three need to be specified, for AVI files all three can be determined from the header.

Exceptions:

invalid_argument if there was an error

Parameters:

fname	the file name of the video
y	the video's rows
x	the video's columns
z	the video's number of frames
colors	the number of colors (always 3 with AVI files)
skip	the size of the header (only raw videos)

void ColorVideo::write	(	char const *	fname,
		int	rawskip = `512`,
		bool	beautify = `false`
	)

Write the video. All steps independent of the file format will be performed, like testing for file writeability etc.

Exceptions:

ios_base::failure	a read error has occured [not supported by all libraries, so eventually {invalid_argument} instead]
invalid_argument	the source video has a different number of colors than the target format allows

Parameters:

fname	the target file name
rawskip	the amount of header area before contents (will be filled with zeroes)
beautify	beautify video that have not had more analysis than synthesis steps?

void ColorVideo::write	(	char const *	fname,
		filetype	ftype,
		int	rawskip = `512`,
		bool	beautify = `false`
	)

Write an video. Writes an video to a file using the specified file type.

Parameters:

ftype	the file type
fname	the file name, if {NULL}, then {stdout}
rawskip	the amount of header area before contents (will be filled with zeroes)
beautify	beautify videos that have not had more analysis than synthesis steps?