The software is designed to evaluate batch of images generated with experimental design of Selgen company. The software is coded in Python3 and mainly uses OpenCV library.
Single image contains two trays (left and right) of 5x8 growing areas. This is region of interest (ROI) for further analysis.
The goal is to evaluate spatial and color pattern of plant in each cell. For experiment "green" pixels are valid. What is "green" can be defined by user in global_variables.py as a thresholds for segmentation.
Software suppose various image formats jpg, png, bmp, tiff, tif as an input data.
The output of analysis is xlsx file described below and folder of raw images with painted trays grid and contours around crop.
Process of single image analysis follows these steps:
- ROI is cropped from raw image
- Tray is splitted into 2 ROI areas (left and right)
- Mask of tray grid is segmented in each ROI
- Computation of ROI mask optimal rotation
- Localization of ROI grid with fourier transform
- ROI separation into growing areas
- Computation of spatial and color pattern in each growing area
- Python virtualenv package is required. Open terminal and execute
python3 -m pip install virtualenvcommand to install this package. - Download zipped selgen project from github repository https://github.com/UPOL-Plant-phenotyping-research-group/Selgen with green button Clone or download and Download ZIP option.
- Open downloaded zip file and place Selgen-master folder on Desktop of your server.
- Open terminal and navigate into project folder with command
cd 'your_path_to_project' (e.g. ...../Desktop/Selgen-master/). - In terminal execute command
bash create_venv, which will create virtual enviroment. Now in project folder ...../Desktop/Selgen-master/ you should find selgen folder.
- As a first python virtualenv package is required. Open command-line interpreter (cmd.exe) terminal and execute
py -m pip install --user virtualenv - Download zipped selgen project from github repository https://github.com/PolakMichalMLT/Selgen with green button Clone or download and Download ZIP option.
- Open downloaded zip file and place Selgen-master folder on Desktop of your server.
- In cmd terminal navigate terminal into this folder
cd .\Desktop\Selgen-master - Create python virtual enviroment with command
python -m venv selgen - Activate virtual enviroment with
.\selgen\Scripts\activate - Install project requirements
pip install -r requirements.txt - Deactivate virtual enviroment with
.\selgen\Scripts\deactivate
- Create folder with images of experiment for processing.
- In project folder ....../Desktop/Selgen-master open file selgen_global.py in some text editor (Notepad++) and define:
- path as directory of folder from step 1
- Navigate terminal to project folder ...../Desktop/Selgen-master.
- In terminal execute command
bash exewhich will execute data analysis.
- Create folder with images of experiment for processing
- In folder ....../Desktop/Selgen-master open file selgen_global.py in some text editor (Notepad++ and define:
- path as directory of folder from step 1
- Navigate cmd terminal to ....../Desktop/Selgen-master
- Activate virtual enviroment with
.\selgen\Scripts\activate - In cmd terminal execute analysis with
python selgen_analysis.py - Deactivate virtual enviroment with
.\selgen\Scripts\deactivate
- all results are located in path from selgen_global.py file
- in contoured_images folder are original images with drawed contours of active biomass
- in processed folder are images, which were sucessfully processed
- in unprocessed folder are images, which were not processed because of som e error
- in batch_output.xlsx are structured results of evaluated batch
- biomass column is value of statistic = segmented pixels of plant in given area
- day column is number of experiment day
- side column specify tray side
- variant column specify treatment/variant
- row column indicates row in tray grid of evaluated area
- column column indicates column in tray grid of evaluated area
- size is size of evaluated area