v1.2.5.3 (20240419)

Data are now stored as JSON files (Plate and Project).
No more use of data_well.txt files.
Using pandas dataframe for Project Overview.
Viewing of selected drop classification over time in Project tab.
Possibility to export to a single csv file all drops with a given classification for further analysis as as spreadsheet.
Added tool to create JSON files for a Plate from data_well.txt files for old projects (./tools/Create_json.py)
Possibility to visualize Human Scores as histogram.
Added option to transfer all screen reservoirs to plate in Tools

Updated GUI.

v1.2.5 (20230113)

Added Tabs to GUI, new Project Tab, updated README, INSTALL
Search_classif_Project.py should work as initially intended.
In the GUI, Search_classif_Project.py is called with the --unique flag. If you don't want this behaviour, you should use the script in the command line.

v1.2.4.5 (20221011)

Fixed core dump if unexpected folder names created by USER in Image_d…

v1.2.4.4 (20220804)

Clicking on a picture in Plate Overview window loads the corresponding data in main window (picture + notes + timeline)
Fixed install script issue with pip
Updated GUI to incorporate screen composition and easy transfer of crystallization cocktail to Notes.
Added possibility to score a drop 1-10. Score is displayed on the heat map.

Updated Setup script.

v1.2.4.3 (20220712)

Clicking on a picture in Plate Overview window loads the corresponding data in main window (picture + notes + timeline)
Fixed install script issue with pip
Updated GUI to incorporate screen composition and easy transfer of crystallization cocktail to Notes

v1.2.4.1 (20211001)

Changed version to 1.2.4.1
Added possibility to use autoMARCO on a single image within the GUI
More efficient load of MARCO model.

v1.2.3.10 (20210610)

Installation procedure with added option regarding folder hierarchy.
Added parameter "_rawimages" in utils.py to edit if necessary before installation.
See INSTALL.txt for more details.

Field showing classification changes color dynamically.
Added preferences file to control shortcuts and cropping routine parameters.
Added requirements.txt for easy install of dependencies with pip.
Can create prep_date.txt within AMI_Image_Analysis.
Added more shortcuts.
Improved navigation with shortcuts: Now only visible wells through "Filter" option are selected.

Added Screens Tables to quickly check crystallization conditions. Crystallization Tables are read from csv files stored in the folder "Screen_Database".

Tool to check circle detection parameters (parameters are stored in preferences.py)
Changed Startup routine, now use file bin/AMI_Image_Analysis.sh.
Added a tool to create bin/AMI_Image_Analysis.sh if using a virtual environment.
Can now save heat map to file.
autoMARCO can work on Raspbian Buster given additional installation steps are performed.
Possibility to save Statistics to csv.

Tensorflow 1.15.2 in requirements.txt (1.15.0 was problematic with Graphics card with CUDA capability <6, needed to use tensorflow 1.14.0)

AutoMARCO grids can all be displayed on screen (previously, only one could be displayed at a time)
Images in the Timeline are clickable and can be displayed in main viewing frame for closer inspection

Merging Z-stack does not need Hugin anymore (unless you want to) and can be done within AMi_Image_Analysis or with the command line

Improved GUI: Improved navigation, Screen composition Tables within GUI

Field showing classification changes color dynamically.
Added preferences file to control shortcuts and cropping routine parameters.
Added requirements.txt for easy install of dependencies with pip.
Can create prep_date.txt within AMI_Image_Analysis.
Added more shortcuts.
Improved navigation with shortcuts: Now only visible wells through "Filter" option are selected.

Added Screens Tables to quickly check crystallization conditions. Crystallization Tables are read from csv files stored in the folder "Screen_Database".

Tool to check circle detection parameters (parameters are stored in preferences.py)
Changed Startup routine, now use file bin/AMI_Image_Analysis.sh.
Added a tool to create bin/AMI_Image_Analysis.sh if using a virtual environment.
Can now save heat map to file.
autoMARCO can work on Raspbian Buster given additional installation steps are performed.
Possibility to save Statistics to csv.

Tensorflow 1.15.2 in requirements.txt (1.15.0 was problematic with Graphics card with CUDA capability <6, needed to use tensorflow 1.14.0)

AutoMARCO grids can all be displayed on screen (previously, only one could be displayed at a time)
Images in the Timeline are clickable and can be displayed in main viewing frame for closer inspection

Merging Z-stack does not need Hugin anymore (unless you want to) and can be done within AMi_Image_Analysis or with the command line

Improved GUI, clickable Timeline

Field showing classification changes color dynamically.
Added preferences file to control shortcuts and cropping routine parameters.
Added requirements.txt for easy install of dependencies with pip.
Can create prep_date.txt within AMI_Image_Analysis.
Added more shortcuts.
Tool to check circle detection parameters (parameters are stored in preferences.py)
Changed Startup routine, now use file bin/AMI_Image_Analysis.sh.
Added a tool to create bin/AMI_Image_Analysis.sh if using a virtual environment.
Can now save heat map to file.
autoMARCO can work on Raspbian Buster given additional installation steps are performed.
Possibility to save Statistics to csv.

Tensorflow 1.15.2 in requirements.txt (1.15.0 was problematic with Graphics card with CUDA capability <6, needed to use tensorflow 1.14.0)

AutoMARCO grids can all be displayed on screen (previously, only one could be displayed at a time)
Images in the Timeline are clickable and can be displayed in main viewing frame for closer inspection

Improved GUI

Field showing classification changes color dynamically.
Added preferences file to control shortcuts and cropping routine parameters.
Added requirements.txt for easy install of dependencies with pip.
Can create prep_date.txt within AMI_Image_Analysis.
Added more shortcuts.
Tool to check circle detection parameters (parameters are stored in preferences.py)
Changed Startup routine, now use file bin/AMI_Image_Analysis.sh.
Added a tool to create bin/AMI_Image_Analysis.sh if using a virtual environment.
Can now save heat map to file.
autoMARCO can work on Raspbian Buster given additional installation steps are performed.
Possibility to save Statistics to csv.

Tensorflow 1.15.2 in requirements.txt (1.15.0 was problematic with Graphics card with CUDA capability <6, needed to use tensorflow 1.14.0)

AutoMARCO grids can all be displayed on screen (previously, only one could be displayed at a time)