Herbarium Sheets

Extract geometric trait measurements from scans or images of herbarium sheets.

Acknowledgements

General approach inspired by Suxing Liu, in particular Smart Plant Growth Top-Down Traits.

PlantCV also used for skeletonization, pruning, leaf & stem counting, etc.

Requirements

Docker is required to run this project in a Unix environment.

Installation

To install from source, clone the project with git clone https://github.com/w-bonelli/herbarium-sheets.git, then build the image from the root directory with docker build -t <your tag> -f Dockerfile ..

Alternatively, you can just pull the pre-built image with docker pull wbonelli/herbarium-sheets, or allow it to be pulled automatically from another Docker CLI command (as below).

Usage

To analyze an image:

docker run wbonelli/herbarium-sheets python3.8 /opt/code/cli.py <input file>

By default, output files will be written to the current working directory. To specify a different output location, use -o <full path to output directory>.

Development

To set up a development environment and explore or modify the source, just mount the project root as your container's working directory, for instance:

docker run -it -v $(pwd):/opt/dev -w /opt/dev wbonelli/herbarium-sheets bash

Then invoke the CLI with python3.8 /opt/code/cli.py <input file>.

Git is also instructed to ignore data and output directories for convenience.