ImageSequence

An image sequence is a collection of image files with sequential filenames.

ImageSequence can be instatiated using:

• a directory name, such as a 'my_directory'

• a “glob” string, such as 'my_directory/*.png', which is safer than using a directory because directories sometimes contain stray files

• the filepath of a zipped archive, such as 'my_directory/all-images.zip'

• a list of filepaths, such as ['image1.png', 'image2.png']

In all cases except the last one, to sort the filenames, sorting keys are generated by splitting the names into digit and non-digit parts (effectively sorting numbers numerically). For example, ['abc9.png', 'abc10.png', 'def9.png'] are in sorted order. If a list of paths is given, it is used as is, with no additional sorting.

ImageSequenceND

See also

The section Multidimensional Readers describes how to deal with multidimensional files.

This class allows reading of N-dimensional Image Sequences with multiple indices in the file name, for instance: ['image_z00_t00.png', 'image_z01_t00.png'].

ReaderSequence

This class allows stacking N-dimensional readers creating another N-dimensional with an extra dimension. For example:

reader = pims.ReaderSequence('path/to/files/image*.tif', pims.BioformatsReader)

Dependencies

Several widely-used Python packages have slightly different implementations of imread, a general purpose image-reading function that understands popular formats like PNG, JPG, TIFF, and others. PIMS requires one of the following packages, in order of decreasing preference.

• scikit-image

• imageio

Scikit-image is installed with the PIMS conda package.