TKernelResampler

Hierarchy

TPersistent

Description

TKernelResampler is a general class for image resampling using arbitrary convolution filters. It implements a fast method for resizing images, by precomputing kernel weights in mapping tables or coefficient bins.

For single-sample look-ups, the GetSample method supports three different kernel modes:

kmDynamic — kernel weights are computed on-the-fly;
kmTableNearest — kernel weights are stored in a table, nearest-neighbor is used for the look-up;
kmTableLinear — same as above, but uses linear interpolation for the look-up.

For further information, see topic about Sampling and Rasterization.

Reference

Properties	Methods	Events
In TKernelResampler:
Kernel	Create
KernelClassName	Destroy
KernelMode	FinalizeSampling
TableSize	GetSample
	PrepareSampling
In TBitmap32Resampler:
Bitmap	Changed
PixelAccessMode	GetSampleBounds
TransformerClass	HasBounds
In TCustomResampler:
Width	Resample
In TNotifiablePersistent:
UpdateCount	BeginUpdate	OnChange
	EndUpdate

TKernelResampler

Hierarchy

Description

Reference

See Also