Skip to contents

Superpixels of a single raster layer

Jakub Nowosad

2024-08-30

Source: vignettes/articles/one_var.Rmd
one_var.Rmd

Superpixels is a collection of segmentation concepts of grouping pixels with similar characteristics. It is often used in computer vision to delineate parts of RGB images that are more meaningful and easier to analyze. When applied to RGB images, each superpixel contains similar colors that also could represent real-world objects. A large number of methods for creating superpixels were developed in the last decades, with the SLIC algorithm (Achanta et al. (2012), doi:10.1109/TPAMI.2012.120) being the most prominent.

The supercells package aims to utilize the concept of superpixels to a variety of spatial data. This package works on spatial data with one variable (e.g., continuous raster), many variables (e.g., RGB rasters), and spatial patterns (e.g., areas in categorical rasters). Therefore, it enables not only to find areas that look similar on an RGB (satellite) image, but also to regionalize areas with comparable values of one or more variables.

This vignette shows how to use the supercells package on a single raster layer. To reproduce the following results on your own computer, install and attach the packages:

library(supercells)    # superpixels for spatial data
library(terra)         # spatial raster data reading and handling
library(sf)            # spatial vector data reading and handling

Default arguments

The first step is to read the input data. Here, we are going to use volcano.tif example data that is already included in the supercells package.

vol = rast(system.file("raster/volcano.tif", package = "supercells"))
plot(vol)

It contains a digital elevation model for the Maunga Whau volcano, where each value represents an elevation above sea level.

The supercells package can be used here to delineate areas with similar elevations. It has one main function, supercells(), which expects at least three arguments:

  • x - an input raster (an object of class SpatRaster from the terra package)
  • k - an expected number of superpixels
  • compactness - a compactness value. Larger values create more compact/even (square) superpixels
vol_slic1 = supercells(vol, k = 50, compactness = 1)
vol_slic1

The supercells() output is an sf object, where each row represents a single superpixel. It stores superpixels ids (supercells), coordinates of the superpixels’ centroids (x and y), and an average of all of the input variables (elevation in this case).

It allows us to plot the superpixels’ borders on top of the input raster data:

plot(vol)
plot(st_geometry(vol_slic1), add = TRUE, lwd = 0.2)

The vol_slic1 object can be a final result of the analysis or used for further grouping or clustering.

Customized arguments

The supercells() function also has a set of optional arguments, including:

  • step - a distance (in the number of cells) between initial superpixel’ centers. It can be used instead of k
  • dist_fun - a distance function used to calculate similarities between raster values
  • clean - should connectivity of the superpixel be enforced? If FALSE, then some superpixels could consist of several nonadjacent regions. It is TRUE by default.
  • iter - a number of iterations performed to create the output. It is 10 by default.

Let’s try to change the values of some of them:

vol_slic2 = supercells(vol, step = 11, compactness = 1, dist_fun = "euclidean", 
                       clean = FALSE, iter = 10)

The output still has the same structure (is an sf object with superpixels ids, coordinates of their centroids, and an average of all of the input variables), but has a slightly different number of superpixels, and their borders are a bit different.

plot(vol)
plot(st_geometry(vol_slic2), add = TRUE, lwd = 0.2)

References

Achanta, Radhakrishna, Appu Shaji, Kevin Smith, Aurelien Lucchi, Pascal Fua, and Sabine Süsstrunk. 2012. “SLIC Superpixels Compared to State-of-the-Art Superpixel Methods.” IEEE Transactions on Pattern Analysis and Machine Intelligence 34 (11): 2274–82.