StarDist 0.7.1

Bugfix release (missing files in stardist-0.7.0.tar.gz).

This only affects (presumably very few) people who can't install from wheels.
Everybody else should already be fine with the 0.7.0 release.

See StarDist 0.7.0 for release notes.

StarDist 0.7.0

With more than one year in the making, this is one of the biggest releases yet:

• Support for Python 3.9
• New support for multi-class object classification (e.g. for classifying nuclei into different types, details below)
• Lots of additions to increase memory/time efficiency (details below)
• Integration functions for the new StarDist napari plugin
• Many bug fixes

Object classification

In addition to detecting all object instances (e.g. cell nuclei), StarDist can now additionally classify each found object instance into a fixed number of different object classes (e.g. cell types). Please see the example notebook for a demonstration with 2D images, but the feature also works for 3D images.

Efficiency

There are many changes (mostly internal) that increase the overall efficiency. Highlights:

• Non-maximum suppression (NMS) has been refactored and is now considerably faster (now uses kd-trees internally, inspired by PR #40 from @GFleishman. Thanks!)
• Better integration of edt to speed up training (we thank @william-silversmith for the discussion in #146)
• CNN prediction is now vastly more memory-efficient for large images (sparse flag, enabled by default)

If that's not memory-efficient enough, there's still the option of block-wise processing for very large images (predict_instances_big).

StarDist 0.6.2

• Provide pre-compiled binaries ("wheels") for most platforms, i.e. no need for compilation during installation
• Bug fixes and improvements
• Improve tests, documentation and notebooks
• Internal changes related to testing

StarDist 0.6.1

Bugfix release, mainly to work around Windows installation issue due to Windows 10 version 2004 fmod bug.

Also fixes a bug that occurred when training was continued for a model after it was loaded from disk. Thanks @Xqua!

StarDist 0.6.0

This is a big release with lots of (internal) changes:

• Support both TensorFlow 1 and 2 (via CSBDeep >= 0.6.0)
• Support prediction for very large images via block-wise processing (cf. #36)
• Remove Python 3.5 support and also add Python 3.8 support (when used with TensorFlow 2)
• Add utility function to export 3D results as .obj file (OBJ geometry format)
• Add possibility to train with data that doesn't fit in memory (cf. #57)
• Fix installation bug when numpy is not installed (#71)
• Host pre-trained models on GitHub (cf. #48)
• Add new matching metrics
• Other bug fixes and improvements

StarDist 0.5.0

Changes:

• Support and include pretrained models (cf. #46)
• Improve ImageJ RoI export (subpixel resolution)
• Fix minor bugs related to polygon representations and plotting
• Deprecate Python 3.5. Starting with StarDist 0.6.0, Python 3.5 will no longer be supported.

Improve example notebooks (although not part of the python package itself):

• Include evaluation and detection performance after training
• Demonstrate how prediction results can be saved (cf. #45)
• Show and use pretrained models
• Fix some minor bugs in the other2D notebooks
• Add data augmentation example for 2D model training
• Notebooks can be run standalone, no need to clone the whole repository anymore

StarDist 0.4.1

Fix installation bug

StarDist 0.4.0

This release has many bug fixes and improvements, including:

• Improve non-maximum suppression (faster, interruptible, verbose option)
• Make 2D distance computation more accurate (issue #33)
• Add export_TF function to export the model for use in the companion ImageJ/Fiji plugin (cf. #10)
• Fix several bugs that only occurred when using multi-channel input images
• Fix some issues with ImageJ ROI export
• Fix bug to allow using input images without annotated objects (issue #25)
• Add plotting functions to show instance segmentation (2D only)
• Add support for building shared library for C++ code (work in progress)
• Fix potential bug when using constant value label images
• Fix potential bug when using the edt package (deal with non-contiguous arrays)

Minor (but potentially breaking) changes:

• Add train_foreground_only configuration parameter
• Data type of output label image now int32 for both 2D and 3D (was uint16 for 2D before)
• Signature of augmenter function changed (now expects a single input/label image pair as input)

StarDist 0.3.6

Minor release

• Better tiling strategy for prediction
• Add tiling parameter in threshold optimization
• Add overlap_label paramter in polyhedron rendering

StarDist 0.3.5

• Small bugfix for windows installation