Changelog

1.4 (26/05/2020)

  • Implemented save quality/lossless options for WebP (@mozgsml)

  • Added missing docs for WebP support (@mozgsml)

1.3 (16/10/2019)

  • Added .get_frame_count() operaton (@kaedroho)

1.2 (11/10/2019)

  • Added WebP support (@frmdstryr)

  • Added .rotate() operaton (@mrchrisadams & @simo97)

1.1 (04/12/2017)

  • Added set_background_color_rgb operation

  • Update MANIFEST.in (Sanny Kumar)

1.0 (04/08/2017)

  • OpenCV 3 support (Will Giddens)

  • Removed Apple copyrighted ICC profile from orientation test images (Christopher Hoskin)

  • Fix: Altered detect_features in OpenCV 3 to return a list instead of a numpy array (Trent Holliday)

  • Support for TIFF files (Maik Hoepfel)

  • Support for BMP files was made official (Maik Hoepfel)

0.4 (05/10/2016)

  • Support for image optimisation and saving progressive JPEG files

  • Added documentation

0.3.1 (16/05/2016)

  • Fixed crash in the Pillow auto_orient operation when the image has an invalid Orientation EXIF Tag (Sigurdur J Eggertsson)

  • The auto_orient operation now catches all errors raised while reading EXIF data (Tomas Olander)

  • Palette formatted PNG and GIF files that have transparency no longer lose their transparency when resizing them

0.3 (09/03/2016)

A major internals refactor has taken place in this release, there are a number of breaking changes:

  • The Image class is now immutable. Previously, “resize” and “crop” operations altered the image in-place but now they now always return a new image leaving the original untouched.

  • There are now multiple Image classes. Each one represents possible state the image can be in (eg, in a file, loaded in Pillow, etc). Operations can return an image in a different class to what the operation was performed on.

  • The “backends” have been renamed to “plugins”.

  • A new registry module has been added which can be used for registering new plugins and operations.

  • The “original_format” attribute has been deprecated.

Other changes in this release:

  • Added auto_orient operation

0.2.1 (27/05/2015)

  • JPEGs are now detected from first two bytes of their file. Allowing non JFIF/EXIF JPEG images to be loaded

0.2 (01/04/2015)

  • Added loader for BMP files

  • Added has_alpha and has_animation operations

  • Added get_pillow_image and get_wand_image operations

  • Added save_as_{jpeg,png,gif} operations

  • Crop and resize now all arguments in a tuple (Similar to Pillow)

  • Dropped Python 2.6 and 3.2 support

  • Formats now detected using images header instead of extension

  • Now possible to specify alternative cascade file for face detection

  • Fix: Images now saved in the same format they were loaded

  • Fix: 1 and P formatted images now converted to RGB when saving to JPEG

0.1 (22/02/2015)

Initial release