Title: | Read and write PNG images |
---|---|
Description: | This package provides an easy and simple way to read, write and display bitmap images stored in the PNG format. It can read and write both files and in-memory raw vectors. |
Authors: | Simon Urbanek <[email protected]> |
Maintainer: | Simon Urbanek <[email protected]> |
License: | GPL-2 | GPL-3 |
Version: | 0.1-8 |
Built: | 2024-11-23 06:42:18 UTC |
Source: | CRAN |
Reads an image from a PNG file/content into a raster array.
readPNG(source, native = FALSE, info = FALSE)
readPNG(source, native = FALSE, info = FALSE)
source |
Either name of the file to read from or a raw vector representing the PNG file content. |
native |
determines the image representation - if |
info |
logical, if |
If native
is FALSE
then an array of the dimensions height
x width x channels. If there is only one channel the result is a
matrix. The values are reals between 0 and 1. If native
is
TRUE
then an object of the class nativeRaster
is
returned instead. The latter cannot be easily computed on but is the
most efficient way to draw using rasterImage
.
Most common files decompress into RGB (3 channels), RGBA (4 channels),
Grayscale (1 channel) or GA (2 channels). Note that G and GA images
cannot be directly used in rasterImage
unless
native
is set to TRUE
because rasterImage
requires
RGB or RGBA format (nativeRaster
is always 8-bit RGBA).
As of png 0.1-2 files with 16-bit channels are converted in full
resolution to the array format, but the nativeRaster
format only
supports 8-bit and therefore a truncation is performed (eight least
significant bits are dropped) with a warning if native
is
TRUE
.
# read a sample file (R logo) img <- readPNG(system.file("img", "Rlogo.png", package="png")) # read it also in native format img.n <- readPNG(system.file("img", "Rlogo.png", package="png"), TRUE) # if your R supports it, we'll plot it if (exists("rasterImage")) { # can plot only in R 2.11.0 and higher plot(1:2, type='n') if (names(dev.cur()) == "windows") { # windows device doesn't support semi-transparency so we'll need # to flatten the image transparent <- img[,,4] == 0 img <- as.raster(img[,,1:3]) img[transparent] <- NA # interpolate must be FALSE on Windows, otherwise R will # try to interpolate transparency and fail rasterImage(img, 1.2, 1.27, 1.8, 1.73, interpolate=FALSE) } else { # any reasonable device will be fine using alpha rasterImage(img, 1.2, 1.27, 1.8, 1.73) rasterImage(img.n, 1.5, 1.5, 1.9, 1.8) } }
# read a sample file (R logo) img <- readPNG(system.file("img", "Rlogo.png", package="png")) # read it also in native format img.n <- readPNG(system.file("img", "Rlogo.png", package="png"), TRUE) # if your R supports it, we'll plot it if (exists("rasterImage")) { # can plot only in R 2.11.0 and higher plot(1:2, type='n') if (names(dev.cur()) == "windows") { # windows device doesn't support semi-transparency so we'll need # to flatten the image transparent <- img[,,4] == 0 img <- as.raster(img[,,1:3]) img[transparent] <- NA # interpolate must be FALSE on Windows, otherwise R will # try to interpolate transparency and fail rasterImage(img, 1.2, 1.27, 1.8, 1.73, interpolate=FALSE) } else { # any reasonable device will be fine using alpha rasterImage(img, 1.2, 1.27, 1.8, 1.73) rasterImage(img.n, 1.5, 1.5, 1.9, 1.8) } }
Create a PNG image from an array or matrix.
writePNG(image, target = raw(), dpi = NULL, asp = NULL, text = NULL, metadata = NULL)
writePNG(image, target = raw(), dpi = NULL, asp = NULL, text = NULL, metadata = NULL)
image |
image represented by a real matrix or array with values in the range of 0 to 1. Values outside this range will be clipped. The object must be either two-dimensional (grayscale matrix) or three dimensional array (third dimension specifying the plane) and must have either one (grayscale), two (grayscale + alpha), three (RGB) or four (RGB + alpha) planes. (For alternative image specifications see details) |
target |
Either name of the file to write, a binary connection or
a raw vector ( |
dpi |
optional, if set, must be a numeric vector of length 1 or 2 specifying the resolution of the image in DPI (dots per inch) for x and y (in that order) - it is recycled to length 2. |
asp |
optional, if set, must be a numeric scalar specifying the
aspect ratio ( |
text |
optional, named character vector of entries that will be
saved in the text chunk of the PNG. Names are used as keys. Note
that the |
metadata |
optional, an R object that will be serialized
into the |
writePNG
takes an image as input and compresses it into PNG
format. The image input is usually a matrix (for grayscale images -
dimensions are width, height) or an array (for color and alpha
images - dimensions are width, height, planes) of reals. The planes
are interpreted in the sequence red, green, blue, alpha.
Alternative representation of an image is of nativeRaster
class
which is an integer matrix with each entry representing one pixel in
binary encoded RGBA format (as used internally by R). It can be
obtained from readPNG
using native = TRUE
.
Finally, writePNG
also supports raw array containing the RGBA
image as bytes. The dimensions of the raw array have to be planes,
width, height (because the storage is interleaved). Currently only 4
planes (RGBA) are supported and the processing is equivalent to that
of a native raster.
The result is either stored in a file (if target
is a file
name), in a raw vector (if target
is a raw vector) or sent to a
binary connection.
If either dpi
or asp
is set, the sPHy
chunk is
generated based on that information. Note that not all image viewers
interpret this setting, and even fewer support non-square pixels.
Either NULL
if the target is a file or a raw vector containing
the compressed PNG image if the target was a raw vector.
Currently writePNG
only produces 8-bit, deflate-compressed,
non-quantized, non-interlaced images. Note in particular that
readPNG
can read 16-bit channels but storing them
back using writePNG
will strip the 8 LSB (irrelevant for
display purposes but possibly relevant for use of PNG in
signal-processing if the input is truly 16-bit wide).
Simon Urbanek
# read a sample file (R logo) img <- readPNG(system.file("img","Rlogo.png",package="png")) # write the image into a raw vector r <- writePNG(img) # read it back again img2 <- readPNG(r) # it better be the same identical(img, img2) # try to write a native raster img3 <- readPNG(system.file("img","Rlogo.png",package="png"), TRUE) r2 <- writePNG(img3) img4 <- readPNG(r2, TRUE) identical(img3, img4) ## text and metadata r <- writePNG(img, text=c(source=R.version.string), metadata=sessionInfo()) img5 <- readPNG(r, info=TRUE) attr(img5, "info") attr(img5, "metadata")
# read a sample file (R logo) img <- readPNG(system.file("img","Rlogo.png",package="png")) # write the image into a raw vector r <- writePNG(img) # read it back again img2 <- readPNG(r) # it better be the same identical(img, img2) # try to write a native raster img3 <- readPNG(system.file("img","Rlogo.png",package="png"), TRUE) r2 <- writePNG(img3) img4 <- readPNG(r2, TRUE) identical(img3, img4) ## text and metadata r <- writePNG(img, text=c(source=R.version.string), metadata=sessionInfo()) img5 <- readPNG(r, info=TRUE) attr(img5, "info") attr(img5, "metadata")