Quantitative dissection of color patterning in the foliar ornamental Coleus reveals underlying features driving aesthetic value

readMe.txt (this file) /Coleus_scans.zip: This directory contains 3 sub-directories:

1. raw_scans: Raw scans of coleus leaves. Data collection described below.

2. binary_leaves: A zip file of individual binary leaves isolated from the raw scans. Data processing described below.

3. colored_leaves: A zip file of individual colored leaves isolated from the raw scans using the binary leaf silhouettes. Data processing described below.

Data collection

Coleus leaf scans were collected from a starting population of 50,000 seedlings that were originally harvested from 133 open-pollinated mother plants in early January in Gainesville, FL. We organized the seedlings into families based on their maternal parents, grew the plants for five weeks and then selected ~2,000 individuals as potential new cultivars based on their foliar color patterning and branching architecture in mid-February. This data represents the youngest fully expanded leaf from each plant between 5-6 weeks of age. Leaves were imaged on Epson Perfection V550 Scanners with Kodak KOCSGS color separation guides included for color calibration. Analysis App Color analysis can be performed using an open-access software program called ColourQuant (Li et al., 2019); available on github: github.com/maoli0923/ColourQuant).

Explanation of isolated binary and colored leaf files

To isolate individual leaves from the raw data scans -  we adjusted the RGB color balance on each scan by a white balance method so that the white swatch in the Kodak KOCSGS color separation guide is pure white, to ensure that scanners were not biasing the color data. Next, we segmented the leaves from the background by converting the RGB matrix into hue-saturation-value (HSV) format. Since most background pixels become grey in HSV, this was used to set a threshold that separates grey values from true leaf values. We then used the binary leaf silhouettes to extract the individual colored leaves by setting the background to pure white, and the foreground to pure black. We manually adjusted the thresholding for leaves that could not be automatically extracted due to shadows in the scan. The binary and colored leaf folder contain outliers, including leaves that were overlapping on the scanner, very small, or broken. These can be manually removed before analysis.  

File ID Key

Files are named with the following code: Year_Family_Scan#. Files containing selected leaves for cultivar development are prepended with an “S” and files containing maternal leaf scans are prepended with an “M.”

For questions regarding released dataset contact: Margaret Frank mhf47@cornell.edu

Li, M., Frank, M.H., and Migicovsky, Z.. 2019a. ColourQuant: A high throughput technique to extract and quantify colour phenotypes from plant images. arXiv 190301652. http://arxiv.org/abs/1903.01652