Netpbmfile is a Python library to read and write image files in the Netpbm or related formats:

• PBM (Portable Bit Map): P1 (text) and P4 (binary)
• PGM (Portable Gray Map): P2 (text) and P5 (binary)
• PPM (Portable Pixel Map): P3 (text) and P6 (binary)
• PNM (Portable Any Map): shorthand for PBM, PGM, and PPM collectively
• PAM (Portable Arbitrary Map): P7, bilevel, gray, and rgb
• PGX (Portable Graymap Signed): PG, signed grayscale
• PFM (Portable Float Map): Pf (gray), PF (rgb), and PF4 (rgba), read-only
• XV thumbnail: P7 332 (rgb332), read-only

The Netpbm formats are specified at https://netpbm.sourceforge.net/doc/.

The PGX format is specified in ITU-T Rec. T.803.

No gamma correction or scaling is performed.

Install the netpbmfile package and all dependencies from the Python Package Index:

python -m pip install -U "netpbmfile[all]"

See Examples for using the programming interface.

Source code and support are available on GitHub.

This revision was tested with the following requirements and dependencies (other versions may work):

• CPython 3.9.13, 3.10.11, 3.11.9, 3.12.3
• NumPy 1.26.4

2024.5.24

• Fix docstring examples not correctly rendered on GitHub.

2024.4.24

• Support NumPy 2.

2023.8.30

• Fix linting issues.
• Add py.typed marker.

2023.6.15

• Drop support for Python 3.8 and numpy < 1.21 (NEP29).
• Improve type hints.

2023.1.1

• Several breaking changes:
• Rename magicnum to magicnumber (breaking).
• Rename tupltypes to tupltype (breaking).
• Change magicnumber and header properties to str (breaking).
• Replace pam parameter with magicnumber (breaking).
• Move byteorder parameter from NetpbmFile.asarray to NetpbmFile (breaking).
• Fix shape and axes properties for multi-image files.
• Add maxval and tupltype parameters to NetpbmFile.fromdata and imwrite.
• Add option to write comment to PNM and PAM files.
• Support writing PGX and text formats.
• Add Google style docstrings.
• Add unittests.

2022.10.25

• …

Refer to the CHANGES file for older revisions.

Write a numpy array to a Netpbm file in grayscale binary format:

>>> data = numpy.array([[0, 1], [65534, 65535]], dtype=numpy.uint16)
>>> imwrite('_tmp.pgm', data)

Read the image data from a Netpbm file as numpy array:

>>> image = imread('_tmp.pgm')
>>> numpy.testing.assert_equal(image, data)

Access meta and image data in a Netpbm file:

>>> with NetpbmFile('_tmp.pgm') as pgm:
...     pgm.magicnumber
...     pgm.axes
...     pgm.shape
...     pgm.dtype
...     pgm.maxval
...     pgm.asarray().tolist()
...
'P5'
'YX'
(2, 2)
dtype('>u2')
65535
[[0, 1], [65534, 65535]]

View the image and metadata in the Netpbm file from the command line:

$ python -m netpbmfile _tmp.pgm