morley 0.0.6.1

GUI software for plant morphometry

Morley

Morley is open-sourse software for plants morphometry: measuring sprouts, roots length, plants area. Morley github repository is here

Overview

Morley is a software tool for measuring morphological parameters of plants using plants photo, which is noninvasive for plants. Morley calculates sprout length, length of the longest root, total length of the roots and plant area, performs statistical analysis and produces plots: bar plots of the parameters of different groups, their comparison in a heatmap and histogramms with distribution of the parameters in different groups.

You can use Morley through the GUI and CLI interfaces by installing the package, or learn the algorithm deeper using the Jupyter Notebook placed here.

Useful links

Read the following information before the first usage.

• Python and Morley installation.
• Protocol for image acquisition. Rules to make plant photos suitable for Morley.
• Extended user guide with explanation of Morley parameters: rotation angle, blurring settings, color ranges.

Installation

Morley requires Python 3.9 or newer. Read how to install Python here. You can install Morley GUI from PyPI:

pip install morley

* use this command in the command line (for example, in Anaconda Prompt)

Alternatively, you can install Morley directly from GitHub:

pip install git+https://github.com/dashabezik/Morley.git

Example

Run morley command on the command line (or in Anaconda Prompt), follow the instructions in the User Guide (more detailed instructions are in the extended user guide). For quickstart you can use example input files. Bigger raw photos datasets, which you can use an advanced example, are placed here.

Check the results in the chosen output folder. An example of the output is available here, results for bigger datasets are published here.

User Guide

Input files

As input, the program takes photographs of plants placed in a row on a plain contrasting background, and an example of a seed photo. Morley supports the most popular image files formats (JPEG, PNG, etc., listed here).

All the photos should be equally oriented (all sprouts on one side, all roots on the other). To convert pixels into millimeters you should locate a size standard near the plants on the spouts side. Read more about how to prepare photos in the protocol for image acquisition.

The folder with photos must have certain structure: head folder contains subfolders with raw photos by groups. Read more here

Search parameters

Before starting the analysis, you must set:

• input directory
• seed template file
• output directory
• angle of photo rotation¹;
• bluring parameters²;
• sprouts and roots color ranges³;
• seeds color range³;
• area of the size standerd, $mm^2$;
• germination threshold, mm;
• indent to calculate plants width ⁴, integer.

1. To choose the needed angle rotate the photo to match the example: sprouts on the left and roots on the right. See the GIF above or check the extended user guide.

2. Bluring parameters are 'morph', 'gauss', 'canny_top'. Example values of these parameters used on real data are listed in Table 1. Check the extended user guide on how to choose the settings for your data. Your goal is to find the values of the parameters such that plant contours are detected accurately. Initial values of the parameters are set, you should just fix them a little bit if needed.

3. Color ranges for sprouts and roots should be chosen as light-green and light-violet colors in HSV system. Color range for seeds should be chosen as their natural color (light-yellow for wheat and peas in HSV system). How to set them read in the user guide. Initial values of the parameters are set, you should just fix them a little bit if needed.

4. The indent from the seed boundary for calculating the width of a plant parts is used to avoid including the seed in the width calculation . The value is a divider for the width of the seed template, the value of the indentation will be the width of the seed divided by the entered number. See recommended values here.