Python 的 Exif 操作库

我正在为 python 寻找好的 exif (EXIF)操作库。与处理速度相比,我更喜欢灵活性(例如,检索提供者专有标记的能力)。你有什么建议?

101418 次浏览

You might want to check out exif-py:

Python library to extract EXIF data from tiff and jpeg files. Very easy to use - $ ./EXIF.py image.jpg

or the Python Imaging Library (PIL):

The Python Imaging Library (PIL) adds image processing capabilities to your Python interpreter. This library supports many file formats, and provides powerful image processing and graphics capabilities.

There's also the aptly named pyexif: http://pyexif.sourceforge.net/

The pyexif python library and tools aims at extracting EXIF information from Jpeg and Tiff files which include it. This information is typically included in images created using digital imaging devices such as digital cameras, digital film scanners, etc.

However, it looks like pyexif hasn't been updated in quite while. They recommend if theirs isn't doing the trick to check out EXIF-py, so you should probably try that one first, as their sourceforge page seems to have some activity there lately, though not much. Finally, using PIL you could do this:

from PIL import Image
from PIL.ExifTags import TAGS


def get_exif(fn):
ret = {}
i = Image.open(fn)
info = i._getexif()
for tag, value in info.items():
decoded = TAGS.get(tag, tag)
ret[decoded] = value
return ret

Disclaimer:
I actually have no idea which is best, this is just what I was able to piece together with Google. :)

I've been using pyexiv2 myself recently, and it seems to fit my needs quite nicely. Perhaps it might suit yours as well.

Note: pyexiv2 is only for python2, for python3 use py3exiv2

You might also look at Gheorghe Milas' jpeg.py library at http://www.emilas.com/jpeg/, which is "A python library to parse, read and write JPEG EXIF, IPTC and COM metadata."

A drawback is that he appears to be hosting his domain on a dynamic IP via DynDNS, so it's not always available.

There are some examples of PIL and EXIF.py usage on ASPN

This article describes a Python module for writing EXIF metadata (and not just reading them) using pure Python. Apparently, none of PIL, pyexif, nor EXIF-py support writing EXIF. pyexiv2 appears to be bleeding-edge and platform-specific.

In Python 2.6 the place of module is different. Use this:

import Image
from ExifTags import TAGS

I started to write my own small library which is based on the code in PIL. check it here.

Exiv2 Based solutions

Exiv2 (exiv2: http://exiv2.org/) is a mature, open-source C++ library that supports reading and writing metadata to many image types (JPEG, PNG, TIFF and many raw formats), understands standard (Xmp, IPTC and Exif) and non-standard metadata ("Makernotes"), and runs on multiple platforms (Windows, Linux, and, with some work, Mac).

Python bindings to exiv2 are:

One advantage of pyexiv2 is that there is a windows build available for python 2.7. A windows build request for gexiv2 is here: https://bugzilla.gnome.org/show_bug.cgi?id=712441

exiv2 and the bindings are all open source (GPL).

somehow i get an attributeError for _getexif with Paolo's code above.. I am using Python 2.6.6 and PIL 1.1.7. Is _getexif obsolete now??

Here's a small modification that worked for me.

from PIL import Image
from PIL.ExifTags import TAGS


def get_exif(fn):
ret = {}
i = Image.open(fn)
#    info = i._getexif()
info = i.tag.tags
for tag, value in info.items():
decoded = TAGS.get(tag, tag)
ret[decoded] = value
return ret

Use PIL :)

import os,sys
from PIL import Image
from PIL.ExifTags import TAGS


if __name__ == '__main__':
for (k,v) in Image.open(sys.argv[1])._getexif().iteritems():
print '%s = %s' % (TAGS.get(k), v)
os.system('pause')

I have been using my own wrappers around http://www.sno.phy.queensu.ca/~phil/exiftool/ -- the reason is that it is very complete, the dev is very active. And not being able to support almost all image formats is a absolute showstopper for the project it is needed for

The drawback of course is that it isn't python, so you would need to use subprocess calls, as I do.

The page at http://redmine.yorba.org/projects/gexiv2/wiki (became https://wiki.gnome.org/Projects/gexiv2) reads now:

This will work equally well with either Python 2 or 3, which makes GExiv2 an excellent replacement for pyexiv2, which only supports Python 2.

So, both Python2 and Python3 are now supported by GExiv2.

Good news.