avir::CImageResizerVars Class Reference

#include <avir.h>

Public Attributes
int	BufLen [2]
int	BufOffs [2]
int	BuildMode
int	elalign
int	ElCount
int	ElCountIO
int	fpalign
int	fppack
double	InGammaMult
bool	IsResize2
double	k
double	o
double	OutGammaMult
double	ox
double	oy
int	packmode
int	ResizeStep
int	RndSeed
CImageResizerThreadPool *	ThreadPool
bool	UseSRGBGamma

Detailed Description

Image resizing variables class.

This is an utility "catch all" class that defines various variables used during image resizing. Several variables that are explicitly initialized in this class' constructor are also used as additional "input" variables to the image resizing function. These variables will not be changed by the avir::CImageResizer<>::resizeImage() function.

Member Data Documentation

int avir::CImageResizerVars::BufOffs[2]

Offsets into the intermediate buffers, used to provide prefix elements required during processing so that no "out of range" access happens. This offset is a multiple of ElCount if pixels are stored in interleaved form.

int avir::CImageResizerVars::BuildMode

The build mode to use, for debugging purposes. Set to -1 to select a minimal-complexity mode automatically. All build modes deliver similar results with minor deviations.

int avir::CImageResizerVars::elalign

Length alignment of arrays of elements. This applies to filters and intermediate buffers: this constant forces filters and scanlines to have a length which is a multiple of this value, for more efficient SIMD implementation.

int avir::CImageResizerVars::ElCount

The number of "fptype" elements used to store 1 pixel.

int avir::CImageResizerVars::ElCountIO

The number of source and destination image's elements used to store 1 pixel.

int avir::CImageResizerVars::fpalign

Suggested alignment size in bytes. This is not a required alignment, because image resizing algorithm cannot be made to have a strictly aligned data access in all cases (e.g. de-interleaved interpolation cannot perform aligned accesses).

int avir::CImageResizerVars::fppack

The number of atomic types stored in a single "fptype" element.

double avir::CImageResizerVars::InGammaMult

Input gamma multiplier, used to convert input data to 0 to 1 range. 0.0 if no gamma is in use.

bool avir::CImageResizerVars::IsResize2

Use optimized "doResize2" function.

double avir::CImageResizerVars::k

Resizing step coefficient, updated to reflect the actually used coefficient during resizing.

double avir::CImageResizerVars::o

Starting pixel offset inside the source image, updated to reflect the actually used offset during resizing.

double avir::CImageResizerVars::OutGammaMult

Output gamma multiplier, used to convert data to 0 to 255/65535 range. 0.0 if no gamma is in use.

double avir::CImageResizerVars::ox

Start X pixel offset within source image (can be negative). Positive offset moves image to the left.

double avir::CImageResizerVars::oy

Start Y pixel offset within source image (can be negative). Positive offset moves image to the top.

int avir::CImageResizerVars::packmode

0 if interleaved packing, 1 if de-interleaved.

int avir::CImageResizerVars::ResizeStep

Index of the resizing step in the latest filtering steps array.

int avir::CImageResizerVars::RndSeed

Random seed parameter. This parameter may be incremented after each random generator initialization. The use of this variable depends on the ditherer implementation.

CImageResizerThreadPool* avir::CImageResizerVars::ThreadPool

Thread pool to be used by the image resizing function. Set to NULL to use single-threaded processing.

bool avir::CImageResizerVars::UseSRGBGamma

Perform sRGB gamma linearization (correction).