Skip to content
forked from carsales/pyheif

Python 3.6+ interface to libheif library

License

Notifications You must be signed in to change notification settings

bnavigator/pyheif

 
 

Repository files navigation

pyheif

Python 3.6+ interface to libheif library using CFFI

Note: currently only reading is supported.

Installation

Simple installation - Linux (installs manylinux2014 wheel, doesn't work with Alpine)

pip install --upgrade pip
pip install pyheif

Installing from source - MacOS

brew install libffi libheif
pip install git+https://github.com/carsales/pyheif.git

Installing from source - Linux

apt install libffi libheif-dev libde265-dev

or

yum install libffi libheif-devel libde265-devel

then

pip install git+https://github.com/carsales/pyheif.git

Installing from source - Windows

Sorry, not going to happen!

Usage

Read the primary image of a HEIF encoded file

The pyheif.read(path_or_bytes) function can be used to read the primary image of a HEIF encoded file. It can be passed any of the following:

  • A string path to a file on disk
  • A pathlib.Path path object
  • A Python bytes or bytearray object containing HEIF content
  • A file-like object with a .read() method that returns bytes

It returns a HeifFile object.

import pyheif

# Using a file path:
heif_file = pyheif.read("IMG_7424.HEIC")
# Or using bytes directly:
heif_file = pyheif.read(open("IMG_7424.HEIC", "rb").read())

Converting to a Pillow Image object

If your HEIF file contains an image that you would like to manipulate, you can do so using the Pillow Python library. You can convert a HeifFile to a Pillow image like so:

from PIL import Image
import pyheif

heif_file = pyheif.read("IMG_7424.HEIC")
image = Image.frombytes(
    heif_file.mode, 
    heif_file.size, 
    heif_file.data,
    "raw",
    heif_file.mode,
    heif_file.stride,
    )

Note: the mode property is passed twice - once to the mode argument of the frombytes method, and again to the mode argument of the raw decoder.

You can now use any Pillow method to manipulate the file. Here's how to convert it to JPEG:

image.save("IMG_7424.jpg", "JPEG")

Read the entire container within the HEIF file

The pyheif.open_container(path_or_bytes) function can be used to read the HEIF container from a HEIF encoded file. It takes the same parameter as pyheif.read()

It returns a HeifContainer object.

Objects

The HeifFile object

The HeifFile has the following properties:

  • mode - the image mode, e.g. "RGB" or "RGBA"
  • size - the size of the image as a (width, height) tuple of integers
  • data - the raw decoded file data, as bytes
  • metadata - a list of metadata dictionaries
  • color_profile - a color profile dictionary
  • stride - the number of bytes in a row of decoded file data
  • bit_depth - the number of bits in each component of a pixel

The UndecodedHeifFile object

This is a HEIF image that has not been decoded. Calling the UndecodedHeifFile.load() method will load the data and the object will become a HeifFile

The HeifContainer object

The HeifContainer has the following properties:

  • primary_image - the HeifTopLevelImage object of the primary image in the file.
  • top_level_images - a list of all HeifTopLevelImage objects in the file.

The HeifTopLevelImage object

The HeifTopLevelImage has the following properties:

  • id - the id of the image
  • image - the UndecodedHeifFile or HeifFile object of the image
  • is_primary - is this the primary image in the container
  • depth_image - the HeifDepthImage if there is one
  • auxiliary_images - a list of HeifAuxiliaryImage objects

The HeifDepthImage object

The HeifDepthImage has the following properties:

  • id - the id of the image
  • image - the UndecodedHeifFile or HeifFile object of the image

The HeifAuxiliaryImage object

The HeifAuxiliaryImage has the following properties:

  • id - the id of the image
  • image - the UndecodedHeifFile or HeifFile object of the image
  • type - a string indicating the type of auxiliary image

About

Python 3.6+ interface to libheif library

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C 64.5%
  • Python 27.2%
  • Dockerfile 7.9%
  • Shell 0.4%