Raster

Classes
struct	MR::RasterInfo
	everything about a raster except its values More...
struct	MR::Raster
	regular grid of values of any type: an image if it has one layer, otherwise a volume More...
struct	MR::RasterLoadSettings
	settings for loading rasters from external formats More...
struct	MR::RasterSaveSettings
	settings for saving rasters in external formats More...

Functions
Expected< Image >	MR::convertRasterToImage (const Raster &raster)
Expected< void >	MR::convertRasterToFloats (const Raster &raster, float *dst, size_t size)
Expected< void >	MR::convertRasterToFloats (const Raster &raster, std::vector< float > &dst)

Detailed Description

Function Documentation

convertRasterToFloats() [1/2]

Expected< void > MR::convertRasterToFloats	(	const Raster &	raster,
		float *	dst,
		size_t	size )

#include <MRMesh/MRRaster.h>

converts the values of a raster to floats in dst, which has room for size = dims.x * dims.y * dims.z values, in the order of the raster:

• one channel is taken as is;
• three or more channels give the luma 0.299 R + 0.587 G + 0.114 B of the first three; fails for two channels, an unknown type, negative dimensions, data whose size does not match the dimensions, or another size

convertRasterToFloats() [2/2]

Expected< void > MR::convertRasterToFloats	(	const Raster &	raster,
		std::vector< float > &	dst )

#include <MRMesh/MRRaster.h>

converts the values of a raster to floats as above, dst is resized to dims.x * dims.y * dims.z values (memory is reallocated only if its capacity is too small)

convertRasterToImage()

Expected< Image > MR::convertRasterToImage

(

const Raster &

raster

)

#include <MRMesh/MRRaster.h>

converts a raster with one layer to an image (the rows are reordered from bottom to top):

• the values of one channel are scaled from the range of finite values to gray colors, NaN and infinite values become black;
• three or more channels are RGB or RGBA colors (the channels after the fourth are ignored): integer values keep their high byte and are taken as unsigned ones, floating-point values are clamped to [0, 1], NaN becomes 0