FreeImage.FillBackground(T) Method

public static bool FillBackground<T>(
	FIBITMAP dib,
	T color,
	FREE_IMAGE_COLOR_OPTIONS options
)
where T : struct, new()

Parameters

dib

Type: FreeImageAPIFIBITMAP
Handle to a FreeImage bitmap.

color

Type: T
The color to fill the bitmap with. See remarks for further details.

options

Type: FreeImageAPIFREE_IMAGE_COLOR_OPTIONS
Options that affect the color search process for palletized images.

Type Parameters

T

The type of the specified color.

Return Value

So, color must be of type Double, if the image to be filled is of type FIT_DOUBLE and must be a FIRGBF structure if the image is of type FIT_RGBF and so on.

However, the fill color is always specified through a RGBQUAD structure for all images of type FIT_BITMAP. So, for 32- and 24-bit images, the red, green and blue members of the RGBQUAD structure are directly used for the image's red, green and blue channel respectively. Although alpha transparent RGBQUAD colors are supported, the alpha channel of a 32-bit image never gets modified by this function. A fill color with an alpha value smaller than 255 gets blended with the image's actual background color, which is determined from the image's bottom-left pixel. So, currently using alpha enabled colors, assumes the image to be unicolor before the fill operation. However, the rgbReserved field is only taken into account, if option [!:FREE_IMAGE_COLOR_OPTIONS.FICO_RGBA] has been specified.

For 16-bit images, the red-, green- and blue components of the specified color are transparently translated into either the 16-bit 555 or 565 representation. This depends on the image's actual red- green- and blue masks.

Special attention must be payed for palletized images. Generally, the RGB color specified is looked up in the image's palette. The found palette index is then used to fill the image. There are some option flags, that affect this lookup process:

Reference