Pure-python AVM library

Project maintained by astrofrog Hosted on GitHub Pages — Theme by mattgraham


PyAVM is a module to represent, read, and write metadata following the Astronomy Visualization Metadata (AVM) standard.


PyAVM supports Python 3.8 and later. No other dependencies are needed simply to read and embed AVM meta-data.

However, the following optional dependencies are needed for more advanced functionality:

Installing and Reporting issues

To install PyAVM, you can simply do:

pip install pyavm

Please report any issues you encounter via the issue tracker on GitHub.

Using PyAVM


PyAVM provides the AVM class to represent AVM meta-data, and is imported as follows:

>>> from pyavm import AVM

Parsing files

To parse AVM meta-data from an existing image, simply call the from_image class method using the filename of the image (or any file-like object):

>>> avm = AVM.from_image('myexample.jpg')

Only JPEG and PNG files are properly supported in that the parsing follows the JPEG and PNG specification. For other file formats, PyAVM will simply scan the contents of the file, looking for an XMP packet. This method is less reliable, but should work in most real-life cases.

Accessing and setting the meta-data

You can view the contents of the AVM object by using

>>> print(avm)

The AVM meta-data can be accessed using the attribute notation:

>>> avm.Spatial.Equinox
>>> avm.Publisher
'Chandra X-ray Observatory'

Tags can be modified:

>>> avm.Spatial.Equinox = "B1950"
>>> avm.Spatial.Notes = "The WCS information was updated on 04/02/2010"

Creating an AVM object from scratch

To create an empty AVM meta-data holder, simply call AVM() without any arguments:

>>> avm = AVM()

Note that this will create an AVM object following the 1.2 specification. If necessary, you can specify which version of the standard to use:

>>> avm = AVM(version=1.1)

Converting to a WCS object

It is possible to create an Astropy WCS object from the AVM meta-data:

>>> wcs = avm.to_wcs()

By default, Spatial.FITSheader will be used if available, but if not, the WCS information is extracted from the other Spatial.* tags. To force PyAVM to not try Spatial.FITSheader, use:

>>> wcs = avm.to_wcs(use_full_header=False)

Initializing from a FITS header

To create an AVM meta-data object from a FITS header, simply pass the header (as an Astropy Header instance) to the from_header class method:

>>> from import fits
>>> header = fits.getheader('image.fits')
>>> avm = AVM.from_header(header)

By default, the AVM tag Spatial.FITSheader will be created, containing the full header (in addition to the other Spatial.* tags). This can be disabled with:

>>> avm = AVM.from_header(header, include_full_header=False)

Initializing from a WCS object

Similarly, it is possible to create an AVM meta-data object from an Astropy WCS instance:

>>> from astropy.wcs import WCS
>>> from pyavm import AVM
>>> wcs = WCS('image.fits')
>>> avm = AVM.from_wcs(wcs)

Tagging images with AVM meta-data

It is possible to embed AVM meta-data into an image file:

>>> avm.embed('original_image.jpg', 'tagged_image.jpg')

At this time, only JPG and PNG files are supported for embedding.