dip::ColorSpaceConverter class

Base class for conversion between two color spaces.

Contents

Classes that convert between color spaces must derive from this and overload all the pure virtual functions. See dip::ColorSpaceManager for how to use these converters.

Functions

auto InputColorSpace() const -> dip::String pure virtual
Returns the source color space name.
auto OutputColorSpace() const -> dip::String pure virtual
Returns the destination color space name.
auto Cost() const -> dip::uint virtual
Returns the cost of the conversion. This cost includes computational cost as well as precision loss.
void Convert(dip::ConstLineIterator& input, dip::LineIterator& output) const pure virtual
This is the method that performs the conversion for one image line.
void SetWhitePoint(dip::XYZ const& whitePoint, dip::XYZMatrix const& matrix, dip::XYZMatrix const& inverseMatrix) virtual
This method is called to set the white point used by the converter. Does nothing by default.

Function documentation

dip::uint Cost() const virtual

Returns the cost of the conversion. This cost includes computational cost as well as precision loss.

Called by dip::ColorSpaceManager::Convert.

The cost is used to avoid pathways such as “RGB” → “grey” → “Lab” instead of “RGB” → “XYZ” → “Yxy” → “Lab”. Conversion to grey therefore must always have a high cost. It is not necessary to define this method, the default implementation returns a cost of 1.

void Convert(dip::ConstLineIterator& input, dip::LineIterator& output) const pure virtual

This is the method that performs the conversion for one image line.

Called by dip::ColorSpaceManager::Convert.

input and output point to buffers with the number of tensor elements expected for the two color spaces, as determined by the InputColorSpace and OutputColorSpace method.

void SetWhitePoint(dip::XYZ const& whitePoint, dip::XYZMatrix const& matrix, dip::XYZMatrix const& inverseMatrix) virtual

This method is called to set the white point used by the converter. Does nothing by default.

Called by dip::ColorSpaceManager::SetWhitePoint.

matrix and inverseMatrix are computed from whitePoint by the caller, to avoid multiple converter functions doing the same computations.