Table of Contents
Research on plant leaf diseases requires the acquisition of quantitative data to characterize the symptoms caused by different pathogens. These symptoms are frequently lesions that are differentiated from the leaf blade by their color and texture. Among the variables used to characterize the impact of a disease, the most relevant are the number of lesions per unit of leaf area, the area and the shape of the lesions. Since visual measurements are only possible on small numbers of images, it is necessary to use computerized image analysis procedures.
Existing procedures can partially meet the needs but are not always adapted to the particularities of the images obtained in the laboratory. From a scanned image under laboratory conditions containing several leaves of plants showing symptoms of a known disease, the algorithm developed makes it possible to obtain for each sheet of the image the number and the characteristics of surface and shape. lesions.
The LeAFtool (Lesion Area Finding tool) is a python script that uses the IPSDK library and also implemented as a macro on the Explorer tool developed by Reactiv'IP.
https://www.reactivip.com/fr/traitement-dimages/#graphic
IPSDK Explorer 3.1; Reactiv’IP; Grenoble, France, 2022; https://www.reactivip.com/
Write to me with your questions.
cd .local/ReactivIP/Explorer/Macro_Interface/
git clone https://github.com/sravel/LeAFtool.gitLeAFtool can be used in command line mode, or GUI macro on Explorer
| Build and Run scripts with interface |
|---|
 |
| Explore results |
|---|
 |
| Live Link to images on Explorer |
|---|
 |
Build config.yaml file like:
log_path:
debug: true
PLANT_MODEL: banana
csv_file:
rename:
RUNSTEP:
draw: true
cut: true
ML: true
merge: true
DRAW-CUT:
images_path:
out_cut_dir:
out_draw_dir:
extension: tif
x_pieces: 1
y_pieces: 1
top: 0
left: 0
right: 0
bottom: 0
noise_remove: false
numbering: Bottom
force_rerun: false
ML:
images_path:
model_name:
model_name_classification:
split_ML: false
calibration_name:
small_object: 30
alpha: 0.8
color_lesion_individual: true
leaf_border: 0
noise_remove: false
force_rerun: false
draw_ML_image: true
MERGE:
rm_original: false
extension: jpgThen run with command line:
~/.local/ReactivIP/Explorer/Macro_Interface/LeAFtool/Leaftool_addons/cmd_LeAFtool.py -c config.yaml- draw (boolean): Active draw tool to see the lines of the border and cutting
- cut (boolean): Active cut tool to see the lines of the border and cutting
- ml (boolean): Active machine learning tool to see apply model build on Explorer
- merge (boolean): Active merge tool to stuck original file to image with classes overlay
- log_path (str): Directory path where log files will be created
- debug (boolean): More verbose logging for errors debug
- PLANT_MODEL (banana/rice): Select plante model
- csv_file (str): CSV path file with Meta-info, scan name and position mandatory at first and second position. Separator is autodetect
- rename (list): ordered list of csv header used to rename cut files
- images_path (str): Input path directory with raw scan images
- out_cut_dir (str): Output path directory for cut images
- out_draw_dir (str): Output path directory for draw images
- extension (jpg/JPG/PNG/png/BMP/bmp/tif/tiff/TIF/TIFF/Tif/Tiff): The raw scan images extension, must be the same for all scan.
- x_pieces (int): The number of output fragments to split vertically Default: 1
- y_pieces (int): The number of output fragments to split horizontally Default: 1
- top (int): The top margin to remove before cut Default: 0
- left (int): The left margin to remove before cut Default: 0
- bottom (int): The bottom margin to remove before cut Default: 0
- right (int): The right margin to remove before cut Default: 0
- noise_remove (boolean): Use IPSDK unionLinearOpening2dImg function to remove small white objet noise Default: False
- force_rerun (boolean): Force running again even files existed, rerun draw and/or cut. Default: False
- numbering (right/bottom): if right: the output order cut is left to right, if bottom: the output order is top to bottom then left Default: right
- images_path (str): Input path directory with cutted scan images
- model_name (int): The IPSDK PixelClassification model name build with Explorer
- model_name_classification (int): The IPSDK Classification model name build with Explorer
- split_ML (boolean): Use machine learning to split leaves instead of RGB Default: False
- calibration_name (str): Name of Explorer calibration, no calibration if empty
- small_object (int): The minimum area of class, to remove small noise detect object Default: 100
- alpha (float): The degree of transparency to apply for overlay color labels. Must float 0 <= alpha <= 1 Default: 0.5
- color_lesion_individual (boolean): If True apply random color for each separated lesions else use all lesions will colored with color of model Default: True
- leaf_border (int): The diameter of the brush (in pixels) used to erode the leaf Default: 0
- noise_remove (boolean): Use IPSDK unionLinearOpening2dImg function to remove small white objet noise Default: False
- force_rerun (boolean): Force running again even files existed. Default: False
- draw_ML_image (boolean): If True, add overlay rectangle corresponding to image used for apply Machine learning (generally one leaf) Default: False
- rm_original (boolean): remove individual files *_mask_overlay Default: False
- extension (jpg/JPG/PNG/png/BMP/bmp/tif/tiff/TIF/TIFF/Tif/Tiff): Merge file extension Default: jpg
| 1 class: lesion, with leaf border |
|---|
 |
| 2 class: lesion and chloroses, without leaf border |
|---|
 |
- Sébastien RAVEL (CIRAD).
Thanks to Cécile Triay for this beautiful logo
Licencied under CeCill-C and GPLv3. Intellectual property belongs to CIRAD and author.