Skip to contents

Overview

Duflor is a package to detect and quantify number of pixels in any subrange(s) of HSV-color space declared by an upper and lower bound. While the package is principally task-agnostic and could work on any sort of HSV-range(s), it is developed for the detection and quantification of plant area.

Installation

You can install the development and release versions of duflor like so:

# release
devtools::install_github("https://github.com/Claudius-Appel/duflor@master")
# development
devtools::install_github("https://github.com/Claudius-Appel/duflor@dev")

If help-vignettes are to be downloaded, use instead:

devtools::install_github("https://github.com/Claudius-Appel/duflor@master"
  ,build_vignettes = T)
devtools::install_github("https://github.com/Claudius-Appel/duflor@dev"
  ,build_vignettes = T)

Vignettes are rendered documents discussing specific topics within a package. Installing them is recommended, but not required for utilising the package itself.

If the front-end shiny-application within the duflor_gui-package is to be used, it is strongly recommended to follow the installation-guide for that package instead.

Prerequisites

Installing this package from GitHub requires the installation of the R-package devtools:

install.packages("devtools")

Install R-build-toolchain

This package contains C++ code. As it is distributed as source-code, users will require the appropriate R-build-toolchain for their respective operating system. For further information and setup-instructions, refer to “https://r-pkgs.org/setup.html#setup-tools”.

For windows, the R-package installr may be used to check if Rtools is already installed, and whether or not it has been found by R. Additionally, it will notify the user if a new R-update is available (major and minor versions only, ignores patch versions):

install.packages("installr") # once pkgbuild is installed, run: 
installr::install.Rtools(check = T,check_r_update = T,GUI = T)

Usage

For detailed documentation, refer to https://claudius-appel.github.io/duflor/

Additionally, various topics have more detailed articles, accessible on the aforementioned website, as well as within R via browseVignettes(). While access from within R is possible, it is generally recommended to view the articles via the website.

A small sidenote on options set by duflor:

Upon loading, duflor will set the following options via options(XX):

  • duflor.default_hsv_spectrums: default HSV-ranges for
    • drought leaf-area
    • green leaf-area
    • complete leaf-area
    • identifier area
  • duflor.default_identifier_area: the size of the identifier_dot in [cm^2].

Their values can be retrieved via getOption(option):

getOption("duflor.default_hsv_spectrums")
getOption("duflor.default_identifier_area")
  • duflor.default_hsv_spectrums is merely used as a shortcut to load the currently-set default hsv ranges. It is not intended to be modified; as it serves no internal use.

  • duflor.default_identifier_area is more likely to require modification. Before calculating the area for each spectrum via duflor::pixels_to_area(pixel.counts), the true area of the identifier-dot must be set via this option. The default is set to 0.503. All areas are assumed to be in [cm^2].