This package contains a set of tools to classify the pixels of digital images into colour categories arbitrarily defined by the user. It is a simple version of the multivariate technique known as Support Vector Machine, adapted to this particular use.
The procedure is simple. A digital image in JPEG or TIFF format is imported into R. The original image contains three colour variables (or bands): (R), (G), and (B). The first step is to transform them into proportions ((r), (g) and (b)), which simplifies the problem into a bivariate one. The pixels of the test images can then be represented in the plane defined by two of the variables (the user judges which two are more convenient by trial and error) and, hopefully, they would form separate clusters (pixel categories). The user then traces straight lines (classification rules) that enclose the pixel clusters. Using the mathematical expression for these rules and the values of the transformed variables, each pixel can be classified in one category. This produces a set of logical matrices (incidence matrices) indicating which pixels belong to each category, stored in appropriate R objects. These can be submitted to posterior analysis or used to create a new version of the original image showing the category of each pixel.
pixelclasser contains functions to visualize the pixels of the images and the rules created by the user, to create the rules and to store them in objects that can be passed to function
classify_pixels() for the analysis of the image, and functions to import and export the original and the classified images.
You can install the last version from the rOpenSci repository in GitHub using packages
devtools, which install
remotes::install_github("ropensci/pixelclasser", build_vignettes = TRUE)
devtools::install_github("ropensci/pixelclasser", build_vignettes = TRUE)
The manual with the description of each function and use examples is the file
/doc/pixelclasser_1.0.0.pdf (see the link to source code on the right), but its contents can be found in the Reference section of this website.
An example session is described in the vignette included in the package, which can be accessed after installation in the usual way:
It also can be accessed in the section Get started in the top menu of this page.
Please note that this package is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.