Bridged Loop Models

A 3D modeling framework for plant phenotyping written in python.

Main Features

• Measure plant morphological traits from photogrammetry-derived 3D surface meshes
• Create a new mesh from an existing mesh where every vertex belongs to an edge loop.
• Define plant structure by the bridges, or connections, between loops
• Compare plants by separating position-based information (bent leaves) from

Additional Features

• Clean up photogrammetry-derived datasets using rule-based approaches to remove unlikely data and add plausible data
• Create procedural models of simulated plants
• Use sunlight simulation models to estimate PAR absorbance and make selections for ideal plant morphology traits

Detailed Features

Mesh to Bridged Loop Model

Convert an existing triangle mesh to a bridged loop model.

1. Slice mesh along multiple directions to extract loop candidates
2. Assess loop candidates and filter out poor-fitting loop candidates
3. Bridge loops by angle and distance

Regenerate mesh from bridged loop model

1. Create triangle mesh from bridged loop model
2. Compare output mesh to input mesh

Plant Trait Extraction from Bridged Loop Model

Extract phenotypic traits from bridged loop model

1. Define segmentation parameters for a specific crop of interest
2. Segment the bridged loop model into plant(s) and plant organs
3. Calculate and summarize traits for all model segments

Getting Started

Installation With Conda:

First-time installation

Clone the repository:

git clone https://github.com/wmpeebles/BridgedLoopModels.git

Create a conda environment

cd BridgedLoopModels/
conda env create -f environment.yml

Activate the environment

conda activate BridgedLoopModels

Install BLooM

pip install -e .

Run the tests

cd tests/
pytest

Updating to the latest version

Update the environment

conda env update -f environment.yml

Update loopmodels

pip install -e .

Uninstall

Remove the conda environment

conda env remove -n BridgedLoopModels