phenotyping-segmentation 0.1.8

A segmentation pipeline for phenotyping

Plant phenotyping based on segmentation model

Purpose and background

Purpose: Segment images and extract traits from segmentation masks for root and shoot images collected from various platforms (cylinder, clearpot, black canvas) with different species (Arabidopsis, rice, soybean, sorghum, maize, pennycress).

Background: To be updated

Installation

You can install by two methods:

1. Clone the repository and navigate to the cloned directory:
   Clone the repository to the local drive.

   git clone https://github.com/Salk-Harnessing-Plants-Initiative/phenotyping-segmentation.git
   cd phenotyping-segmentation
   

2. Install a PyPI package:

   pip install phenotyping-segmentation
   

Organize the pipeline and your images

Models can be downloaded from Box.

Please make sure to organize the downloaded pipeline, model, and your own images in the following architecture:

phenotyping-segmentation/
├── images/
│   ├── wave name (e.g., wave1)/
│   │   ├── day name (e.g., day7)/
│   │   │   ├── plant name (e.g., ZHOKUWVOIZ)/
│   │   │   │   ├── frame image (e.g., 1.png)
├── scans.csv (the image path and scanner id information)
├── model name (e.g., arabidopsis_model.pth)
├── label_class_dict_lr.csv (class color)
├── params.json (pipeline parameter json file)
├── env.yaml (environment file)
├── Dockerfile
├── pipeline.sh (indicate input_dir and pipeline name)

Running the pipeline with a shell file (pipeline.sh)

1. create the environment: In terminal, navigate to your root folder and type:

   conda env create -f env.yaml
   

   or

   mamba env create -f env.yaml
   

2. activate the environment:

   conda activate phenotyping-segmentation
   

3. run the shell file:

   sed -i 's/\r$//' pipeline.sh
   bash pipeline.sh
   

Running the pipeline with a pip installed

1. activate your environment:

   conda activate your-environment-name
   

2. install the pip package:

   pip install phenotyping-segmentation
   

3. run the shell file:

   sed -i 's/\r$//' pipeline.sh
   bash pipeline.sh
   

Running the pipeline with docker

Make sure you have images, and associated files listed above in your root folder.

1. build the docker:

   docker build -t phenoseg .
   

2. run the docker:

   docker run --gpus all phenoseg