Using libvips from C [src]
Using libvips from C
libvips comes with a convenient, high-level C API. You should read the API docs for full details, but this section will try to give a brief overview.
Library startup
When your program starts, use VIPS_INIT()
to start up libvips. You
should pass it the name of your program, usually argv[0]
. If
you call vips_shutdown()
just before you exit, libvips will attempt to free
all resources. This can help leak checking, but is not required.
VIPS_INIT()
is a macro to let it check that the libvips shared library you
have linked to matches the libvips headers you included.
You can add the libvips flags to your GObject
command-line
processing with vips_add_option_entries()
.
The VipsImage
class
The basic data object is the VipsImage
. You can create an image from a
file on disc or from an area of memory, either as a C-style array, or as a
formatted object, like JPEG. See vips_image_new_from_file()
and friends.
Loading an image is fast: libvips read just enough of the image to be able
to get the various properties, such as width, but no decoding occurs until
pixel values are really needed.
Once you have an image, you can get properties from it in the usual
way. You can use projection functions, like vips_image_get_width()
or
g_object_get()
, to get GObject
properties. All
libvips objects are immutable, meaning you can only get properties, you
can’t set them. See libvips Header to read about image properties.
Reference counting
libvips is based on the GObject
library and is therefore
reference counted. vips_image_new_from_file()
returns an object with a count
of 1. When you are done with an image, use g_object_unref()
to
dispose of it. If you pass an image to an operation and that operation needs
to keep a copy of the image, it will ref it. So you can unref an image as soon
as you no longer need it, you don’t need to hang on to it in case anyone else
is still using it.
See VipsOperation
for more details on libvips’ reference counting
conventions. See the Reference pools section below
for a way to automate reference counting in C.
libvips operations
Use things like vips_embed()
to manipulate your images. You use it
from C like this:
const char *filename;
VipsImage *in = vips_image_new_from_file(filename, NULL);
const int x = 10;
const int y = 10;
const int width = 1000;
const int height = 1000;
VipsImage *out;
if (vips_embed(in, &out, x, y, width, height, NULL))
error_handling();
Now out
will hold a reference to a 1000 by 1000 pixel image, with in
pasted 10 right and 10 down from the top left-hand corner. The remainder
of the image will be black. If in
is too large, it will be clipped at
the image edges.
Operations can take optional arguments. You give these as a set of NULL-terminated name-value pairs at the end of the call. For example, you can write:
if (vips_embed(in, &out, x, y, width, height,
"extend", VIPS_EXTEND_COPY,
NULL))
error_handling();
And now the new edge pixels, which were black, will be filled with a copy
of the edge pixels of in
. Operation options are listed at the top of each
operation’s entry in the docs. Alternatively, the vips
program is handy
for getting a summary of an operation’s parameters. For example:
$ vips embed
embed an image in a larger image
usage:
embed in out x y width height [--option-name option-value ...]
where:
in - Input image, input VipsImage
out - Output image, output VipsImage
x - Left edge of input in output, input gint
default: 0
min: -1000000000, max: 1000000000
y - Top edge of input in output, input gint
default: 0
min: -1000000000, max: 1000000000
width - Image width in pixels, input gint
default: 1
min: 1, max: 1000000000
height - Image height in pixels, input gint
default: 1
min: 1, max: 1000000000
optional arguments:
extend - How to generate the extra pixels, input VipsExtend
default enum: black
allowed enums: black, copy, repeat, mirror, white, background
background - Colour for background pixels, input VipsArrayDouble
See VipsOperation
for more information on running operations on images.
The API docs have a handy table of all libvips operations, if you want to find out how to do something, try searching that.
When you are done, you can write the final image to a disc file,
to a formatted memory buffer, or to C-style memory array. See
vips_image_write_to_file()
and friends.
Getting pixels
Use VipsRegion
to read pixels out of images. You can use VIPS_IMAGE_ADDR()
as well, but this can need a large amount of memory to work. See extending for an introduction to writing your own operations.
Error handling
libvips keeps a log of error message, see vips_error_buffer()
and
vips_error_clear()
to find out how to get and clear the error log.
Example
On Unix systems, you can compile the example code with something like:
$ gcc -g -Wall myprog.c `pkg-config vips --cflags --libs`
On Windows, you’ll need to set the compiler flags by hand, perhaps:
x86_64-w64-mingw32-gcc-win32 -mms-bitfields \
-Ic:/vips-8.6/include \
-Ic:/vips-8.6/include/glib-2.0 \
-Ic:/vips-8.6/lib/glib-2.0/include \
myprog.c \
-Lc:/vips-8.6/lib \
-lvips -lz -ljpeg -lstdc++ -lxml2 -lfftw3 -lm -lMagickWand -llcms2 \
-lopenslide -lcfitsio -lpangoft2-1.0 -ltiff -lpng14 -lexif \
-lMagickCore -lpango-1.0 -lfreetype -lfontconfig -lgobject-2.0 \
-lgmodule-2.0 -lgthread-2.0 -lglib-2.0 -lintl \
-o myprog.exe
libvips from C example
#include <stdio.h>
#include <vips/vips.h>
int
main(int argc, char **argv)
{
VipsImage *in;
double mean;
VipsImage *out;
if (VIPS_INIT(argv[0]))
vips_error_exit(NULL);
if (argc != 3)
vips_error_exit("usage: %s infile outfile", argv[0]);
if (!(in = vips_image_new_from_file(argv[1], NULL)))
vips_error_exit(NULL);
printf("image width = %d\n", vips_image_get_width(in));
if (vips_avg(in, &mean, NULL))
vips_error_exit(NULL);
printf("mean pixel value = %g\n", mean);
if (vips_invert(in, &out, NULL))
vips_error_exit(NULL);
g_object_unref(in);
if (vips_image_write_to_file(out, argv[2], NULL))
vips_error_exit(NULL);
g_object_unref(out);
return 0;
}
Reference pools
libvips has a simple system to automate at least some reference counting issues. Reference pools are arrays of object pointers which will be released automatically when some other object is finalized.
The code below crops a many-page image (perhaps a GIF or PDF). It splits the
image into separate pages, crops each page, reassembles the cropped areas,
and saves again. It creates a context
object representing the state of
processing, and crop_animation
allocates two reference pools off that
using vips_object_local_array()
, one to hold the cropped frames, and one
to assemble and copy the result.
All unreffing is handled by main
, and it doesn’t need to know anything
about crop_animation
.
#include <vips/vips.h>
static int
crop_animation(VipsObject *context, VipsImage *image, VipsImage **out,
int left, int top, int width, int height)
{
int page_height = vips_image_get_page_height(image);
int n_pages = image->Ysize / page_height;
VipsImage **page = (VipsImage **) vips_object_local_array(context, n_pages);
VipsImage **copy = (VipsImage **) vips_object_local_array(context, 1);
int i;
/* Split the image into cropped frames.
*/
for (i = 0; i < n_pages; i++)
if (vips_crop(image, &page[i],
left, page_height * i + top, width, height, NULL))
return -1;
/* Reassemble the frames and set the page height. You must copy before
* modifying metadata.
*/
if (vips_arrayjoin(page, ©[0], n_pages, "across", 1, NULL) ||
vips_copy(copy[0], out, NULL))
return -1;
vips_image_set_int(*out, "page-height", height);
return 0;
}
int
main(int argc, char **argv)
{
VipsImage *image;
VipsObject *context;
VipsImage *x;
if (VIPS_INIT(NULL))
vips_error_exit(NULL);
if (!(image = vips_image_new_from_file(argv[1],
"access", VIPS_ACCESS_SEQUENTIAL,
NULL)))
vips_error_exit(NULL);
context = VIPS_OBJECT(<a href="ctor.Image.new.html"><code>vips_image_new()</code></a>);
if (crop_animation(context, image, &x, 10, 10, 500, 500)) {
g_object_unref(image);
g_object_unref(context);
vips_error_exit(NULL);
}
g_object_unref(image);
g_object_unref(context);
image = x;
if (vips_image_write_to_file(image, argv[2], NULL)) {
g_object_unref(image);
vips_error_exit(NULL);
}
g_object_unref(image);
return 0;
}