Purpose: Segment images and extract traits from segmentation masks for clearpot images with different species (canola and soybean).
Background: To be updated with manuscript
-
Clone the repository:
git clone https://github.com/Salk-Harnessing-Plants-Initiative/ClearDepth-WRT.git -
Navigate to the cloned directory:
cd ClearDepth-WRT
Models can be downloaded from Box.
Please make sure to organize the downloaded pipeline, model, and your own images in the following architecture:
ClearDepth-WRT/
├── images/
│ ├── experimental design (e.g., day7, or batchA)/
│ │ ├── image (e.g., E-102_1.PNG)/
├── model/
│ ├── label_class_dict_lr.csv (class color)
│ ├── model name
├── env.yaml (environment file)