KoConvolutionOp Class Reference

#include <KoConvolutionOp.h>

Inheritance diagram for KoConvolutionOp:

Public Member Functions
virtual void	convolveColors (const quint8 *const *colors, const qreal *kernelValues, quint8 *dst, qreal factor, qreal offset, qint32 nColors, const QBitArray &channelFlags) const =0
virtual	~KoConvolutionOp ()

Detailed Description

Base class of a convolution operation. A convolution operation is defined by sum(colors[i] * kernelValues[i]) / factor + offset). The most well known convolution is the gaussian blur.

You access the KoConvolutionOp of a colorspace by calling KoColorSpace::convolutionOp.

Definition at line 18 of file KoConvolutionOp.h.

Constructor & Destructor Documentation

~KoConvolutionOp()

virtual KoConvolutionOp::~KoConvolutionOp ( )

inlinevirtual

Definition at line 21 of file KoConvolutionOp.h.

{ }

Member Function Documentation

convolveColors()

virtual void KoConvolutionOp::convolveColors ( const quint8 *const * colors, const qreal * kernelValues, quint8 * dst, qreal factor, qreal offset, qint32 nColors, const QBitArray & channelFlags ) const

pure virtual

Convolve the colors.

Parameters

colors	a pointer toward the source pixels
kernelValues	the coefficient of the source pixels
dst	the destination pixel
factor	usually the factor is equal to the sum of kernelValues
offset	the offset which is added to the result, useful when the sum of kernelValues is equal to 0
nColors	the number of pixels in the colors array
channelFlags	determines which channels are affected in pixel order

This function is thread-safe.

Implemented in KoConvolutionOpImpl< _CSTrait >.

---

The documentation for this class was generated from the following file:

• libs/pigment/KoConvolutionOp.h