pylibjpeg-openjpeg 2.1.0

A Python wrapper for openjpeg, with a focus on use as a plugin for for pylibjpeg

pylibjpeg-openjpeg

A Python 3.8+ wrapper for openjpeg, with a focus on use as a plugin for pylibjpeg.

Linux, OSX and Windows are all supported.

Installation

Dependencies

NumPy

Installing the current release

python -m pip install -U pylibjpeg-openjpeg

Installing the development version

Make sure Python, Git and CMake are installed. For Windows, you also need to install Microsoft's C++ Build Tools.

git clone --recurse-submodules https://github.com/pydicom/pylibjpeg-openjpeg
python -m pip install pylibjpeg-openjpeg

Supported JPEG Formats

Decoding

ISO/IEC Standard	ITU Equivalent	JPEG Format
15444-1	T.800	JPEG 2000

Encoding

Encoding of NumPy ndarrays is supported for the following:

• Array dtype: bool, uint8, int8, uint16, int16, uint32 and int32 (1-24 bit-depth only)
• Array shape: (rows, columns) and (rows, columns, planes)
• Number of rows/columns: up to 65535
• Number of planes: 1, 3 or 4

Transfer Syntaxes

UID	Description
1.2.840.10008.1.2.4.90	JPEG 2000 Image Compression (Lossless Only)
1.2.840.10008.1.2.4.91	JPEG 2000 Image Compression
1.2.840.10008.1.2.4.201	High-Throughput JPEG 2000 Image Compression (Lossless Only)
1.2.840.10008.1.2.4.202	High-Throughput JPEG 2000 with RPCL Options Image Compression (Lossless Only)
1.2.840.10008.1.2.4.203	High-Throughput JPEG 2000 Image Compression

Usage

With pylibjpeg and pydicom

from pydicom import dcmread
from pydicom.data import get_testdata_file

ds = dcmread(get_testdata_file('JPEG2000.dcm'))
arr = ds.pixel_array

Standalone JPEG decoding

You can also decode JPEG 2000 images to a numpy ndarray:

from openjpeg import decode

with open('filename.j2k', 'rb') as f:
    # Returns a numpy array
    arr = decode(f)

# Or simply...
arr = decode('filename.j2k')

Standalone JPEG encoding

Lossless encoding of RGB with multiple-component transformation:

import numpy as np
from openjpeg import encode

arr = np.random.randint(low=0, high=65535, size=(100, 100, 3), dtype="uint8")
encode(arr, photometric_interpretation=1)  # 1: sRGB

Lossy encoding of a monochrome image using compression ratios:

import numpy as np
from openjpeg import encode

arr = np.random.randint(low=-2**15, high=2**15 - 1, size=(100, 100), dtype="int8")
# You must determine your own values for `compression_ratios`
#   as these are for illustration purposes only
encode(arr, compression_ratios=[2, 4, 6])

Lossy encoding of a monochrome image using peak signal-to-noise ratios:

import numpy as np
from openjpeg import encode

arr = np.random.randint(low=-2**15, high=2**15 - 1, size=(100, 100), dtype="int8")
# You must determine your own values for `signal_noise_ratios`
#   as these are for illustration purposes only
encode(arr, signal_noise_ratios=[50, 80, 100])

See the docstring for the encode() function for full details.