The following major changes were introduced with CodraFT V2:
* Fully automated high-level processing features for internal testing purpose, as well
as embedding CodraFT in a third-party software
* Extensive test suite (unit tests and application tests) with 90% feature coverage
* Segmentation fault and Python exception logging
* Customizable annotations for both signals and images
Release key features
* New data visualization and processing features:
| Signal | Image | Feature |
|:------:|:-----:|--------------------------------------------------------|
| | • | Automatic 2D-peak detection |
| | • | Automatic contour extraction (circle/ellipse fit) |
| • | • | Multiple Regions of Interest (ROIs) |
| | • | User-defined annotations (labels and geometric shapes) |
| • | • | "Statistics" computing feature |
* Automation of high-level processing features: added fully automated high-level test
scenarios, and enhanced public API for embedding CodraFT into a third-party application
* Test Driven Development with high quality standards
(pylint score >= 9.8/10, test coverage >= 90%)
Detailed feature list
New data visualization and processing features:
* Image:
* New automatic image contour detection feature returning fitted circle/ellipse
* New automatic 2D peak detection feature (optionally create ROIs)
* "View in a new window": added customizable "Annotations" support for both signal
and image panels - supports user-defined annotations (points, segments, circles,
ellipses, labels,...) which are serialized in image metadata
* Added "Show graphical object titles" option in "View" menu to show or hide the title
(or subtitle) of ROIs or any other graphical object
* Added support for **multiple** Regions of Interest (ROI):
* All "Computing" menu features apply to multiple ROIs
* Computation result arrays now contains ROI index (first column) and one row per ROI
* ROI are merged when summing objects (signals or images)
* ROI can be removed, modified or added at any time
* Added option "Show graphical object titles" ("View" menu) to show or hide ROI titles or any
other geometrical shapes title (or subtitle)
* New computing "Statistics" feature showing a table with statistics on image/signal
and eventually regions of interest (min, max, mean, standard deviation, sum, ...)
New general purpose features:
* Memory management:
* New available memory indicator on main window status bar
* New warning dialog box when trying to open/create data if available memory is below
the "available_memory_threshold" defined in CodraFT configuration file (default: 500MB)
* Error handling:
* New integrated log file viewer
* New warning dialog box at startup suggesting to view log files when logs were
generated during last session
* Logging segmentation faults in ".CodraFT_faulthandler.log"
* Logging Python exceptions in ".CodraFT_traceback.log"
* Signal/Image metadata:
* New copy/paste feature: update object metadata from another one
* New import/export feature: import-export object metadata (JSON text file) using the
new "Import metadata into" / "Export metadata from" entries in "File" menu
* HDF5 browser feature: complete redesign (better compatibility, evolutive design, ...)
* Added support for multiple HDF5 files opening at once
* Added `.CodraFT.ini` configuration file (user home directory):
* New configuration file entry: current working directory
* New configuration file entry: current main window size and position
* New configuration file entry: embedded Python console enable state
* New configuration file entry: available memory alarm threshold
New test-related features:
* Added non-interactive tests, opening the way for unit tests with better coverage
* Added "unattended" and "screenshot" execution modes respectively for testing and documentation purpose
* Added automated high-level test scenarios (signal and image processing)
* Tests are now splitted in two categories: unit tests (`*_unit.py`) and application tests (`*_app.py`).
* Added Coverage.py support
* Added "all_tests.py" to run all tests in unattended mode
New dependencies:
* [scikit-image](https://pypi.org/project/scikit-image/)
* [psutil](https://pypi.org/project/psutil/)
Other changes (on existing features):
* Image and Signal:
* Object properties panel: added data type information (feature refactored upstream to guidata)
* New random signal/image: added support for both Normal and Uniform distributions
* Operations "sum" and "average" now merge metadata results
* Computed titles "s/i000" are now renamed after inserting/removing an object
* Computing results (geometrical shapes: segment, circle, ellipse): numerical results
are now automatically added to metadata (respectively: length, center and radius,
center, a and b)
* Image:
* Added support for image origin and pixel size
* Flat field correction: added threshold parameter
* "New image" now creates an image with the same data type as selected image
* "New image" now supports uint16 data type
* Signal:
* Peak detection: added minimal distance parameter
* Fit dialog / plot: do auto scale at startup
* Peak detection dialog: preselect horizontal cursor at startup
* `codraft.core.gui` code refactoring: added subpackage `core.gui.processor`
* Added "Browse HDF5" action to main window ("Open HDF5" now imports all data)
Bug fixes:
* HDF5 file import: converted `bytes` metadata to `str`
* Added h5py to requirements (setup.py)
* Plot: reintroduced pure white background in light mode (white background was removed
unintentionally when introducing dark mode)
* Image:
* "Clean-up data view" feature was accidently removing grid
* Fixed hard crash when trying to visualize images with NaNs (use case: result of
any filter on `uint8` image)
* Fixed hard crash when using image Z-axis log scale on some images
* Fixed DICOM support
* Fixed hard crash in "to_codraft" (cross section item with empty data)
* Fixed image visualization parameters update from metadata
* MinEnclosingCircle: fixed sqrt(2) error
* Signal:
* "Clean-up data view" feature was accidently removing legend box and grid
* Fixed integral (missing initial point)
* Fixed plotting support for complex data
* Fixed signal visualization parameters update from metadata