Declaration

TVectorMap = class(TCustomMap)

Description

TVectorMap is a 2D map container that stores coordinate displacement vectors (TFixedVector / TFloatVector) for each coordinate in a rectangular grid. It is primarily used in Graphics32 to perform non-linear spatial image warping, resampler distortions, and mesh deformations (such as in TRemapTransformation).

Purpose

In spatial transformations, a vector map represents an offset field $(\mathrm{\Delta }X,\mathrm{\Delta }Y)$ for every pixel in a destination surface. When applying a vector map transformation to an image, the sampler inspects the displacement vector at $(X,Y)$ and samples the source image from $(X+\mathrm{\Delta }X,Y+\mathrm{\Delta }Y)$.

Typical Population Methods

TVectorMap instances are typically populated in one of the following ways:

1. Rasterizing a Transformation: Using RasterizeTransformation (from GR32_Transforms) to evaluate a TTransformation object across a destination rectangle and record generated displacement vectors.
2. Procedural / Mathematical Vector Generation: Writing displacement vectors directly via indexed pixel accessor properties (FixedVector, FloatVector, FixedVectorX, FloatVectorF).
3. Loading Mesh Files: Importing Adobe Photoshop Liquify mesh files (.msh) using TVectorMap.LoadFromFile.
4. Vector Map Merging: Combining or layering existing vector map displacement fields using TVectorMap.Merge.