Plantcv

[](https://doi.org/10.5281/zenodo.2532362)

PlantCV v3.0.3 fixes some minor issues with PyPI deployment.

[](https://doi.org/10.5281/zenodo.2532116)

PlantCV v3.0.2 only changes PyPI authentication information.

[](https://doi.org/10.5281/zenodo.2531970)

PlantCV v3.0.1 adds a new Travis CI build method on git tags that should also trigger auto-deployment to PyPI.

* `plantcv.fill` is now significantly faster using a scikit-image backend
* `plantcv.median_blur` now allows the kernel size to be asymmetric (a two-element tuple)
* Fixed issues with Read the Docs rendering
* Documentation was updated to fix typos, command examples, etc.
* Some test data was reduced in size to shrink the memory footprint during build tests
* Fixed some package usage syntax to avoid future method deprecation issues
* Added color card mask helper function
* Added color card checker analysis function
* Added color card auto detection method (thanks to bhive01 for the method inspiration)
* Improved code testing coverage to 98%
* Added method to read raw Bayer images
* Updated `plantcv.fluor_fvfm` to accept 8-bit images in addition to 16-bit images
* Added conda environment YAML configuration files for quick conda installation

PlantCV v3.0.dev2 is an early release in the v3 release cycle. A full v3 will be released at a later date. While this is an early release, it should be considered essentially production-ready. Note that this release breaks previous syntax/functionality with the goal of making things easier. Updating existing code to the new styles should be fairly straightforward.

[](https://doi.org/10.5281/zenodo.1408271)

Overview of changes:

* The input and output `device` has been eliminated for all functions. The step counter still exists in the background.
* The input `debug` has been eliminated for all functions. It can now be set globally with the class `Params`: https://plantcv.readthedocs.io/en/latest/params/.
* The output location of images printed in debug mode can now be set using the `plantcv.params.debug_outdir` global setting: https://plantcv.readthedocs.io/en/latest/params/
* Some function arguments were refactored. Many were changed to make it more explicit what type of input image is required. The new vocabulary is generally: `img` = all image types, `rgb_img` = RGB images only, `gray_img` = Grayscale images only, `bin_img` or `mask` = Binary images only. Some other inputs were renamed for consistency and/or PEP8 formatting guidelines. All function call changes are summarized here: https://plantcv.readthedocs.io/en/latest/updating/#updating-to-v3.
* Several functions were removed and replaced with updated versions.
* `define_roi`: replaced with the functions in the subpackage `roi`.
* `analyze_bound`: replaced with the functions `analyze_bound_horizontal` and `analyze_bound_vertical`.
* `rotate_img`: replaced with the function `rotate`.
* The threshold functions `binary_threshold`, `adaptive_threshold`, `otsu_auto_threshold`, and `triangle_auto_threshold` were moved to a new subpackage and are now: `plantcv.threshold.binary`, `plantcv.threshold.gaussian`, `plantcv.threshold.mean`, `plantcv.threshold.otsu`, `plantcv.threshold.triangle`.
* These changes are also summarized here: https://plantcv.readthedocs.io/en/latest/updating/#updating-to-v3.
* Testing coverage was increased to 94%.
* Some additional general documentation and docstring reformatting for consistency.
* The FAQ was updated.

[](https://doi.org/10.5281/zenodo.1296371)

Some features already included in the v3 release:

* Distance transformation
* Release of PlantCV through Docker containerization
* Full and tested Python2/3 and OpenCV2/3 compatibility
* Global settings and simplified API framework
* New boundary line tools
* Simplified region of interest (ROI) tools
* Color correction/standardization
* Package namespace reorganization
* Documentation improvements and bug fixes