unsupervised classification with R

unsupervised classification with R

m

January 29, 2016

Here we see three simple ways to perform an unsupervised classification on a raster dataset in R. I will show these approaches, but first we need to load the relevant packages and the actual data. You could use the Landsat data used in the “Remote Sensing and GIS for Ecologists” book which can be downloaded here.

library("raster")  
library("cluster")
library("randomForest")

# loading the layerstack  
# here we use a subset of the Landsat dataset from "Remote Sensing and GIS for Ecologists" 
image <- stack("path/to/raster")
plotRGB(image, r=3,g=2,b=1,stretch="hist")

RGBimage

Now we will prepare the data for the classifications. First we convert the raster data in a matrix, then we remove the NA-values.

## returns the values of the raster dataset and write them in a matrix. 
v <- getValues(image)
i <- which(!is.na(v))
v <- na.omit(v)

The first classification method is the well-known k-means method. It separates n observations into  k clusters. Each observation belongs to the cluster with the nearest mean.

## kmeans classification 
E <- kmeans(v, 12, iter.max = 100, nstart = 10)
kmeans_raster <- raster(image)
kmeans_raster[i] <- E$cluster
plot(kmeans_raster)

Kmeans

The second classification method is called clara (Clustering for Large Applications). It work by clustering only a sample of the dataset and then assigns all object in the dataset to the clusters.

## clara classification 
clus <- clara(v,12,samples=500,metric="manhattan",pamLike=T)
clara_raster <- raster(image)
clara_raster[i] <- clus$clustering
plot(clara_raster)

clara

The third method uses a random Forest model to calculate proximity values. These values were clustered using k-means. The clusters are used to train another random Forest model for classification.

## unsupervised randomForest classification using kmeans
vx<-v[sample(nrow(v), 500),]
rf = randomForest(vx)
rf_prox <- randomForest(vx,ntree = 1000, proximity = TRUE)$proximity

E_rf <- kmeans(rf_prox, 12, iter.max = 100, nstart = 10)
rf <- randomForest(vx,as.factor(E_rf$cluster),ntree = 500)
rf_raster<- predict(image,rf)
plot(rf_raster)

randomForest

The three classifications are stacked into one layerstack and plotted for comparison.

class_stack <- stack(kmeans_raster,clara_raster,rf_raster)
names(class_stack) <- c("kmeans","clara","randomForest")

plot(class_stack)

Comparing the three classifications:

Looking at the different classifications we notice, that the kmeans and clara classifications have only minor differences.
The randomForest classification shows a different image.

 

want to read more about R and classifications? check out this book:

you may also like:

Course on Object-based image analysis

Course on Object-based image analysis

Dr. Michael Wurm from the German Aerospace Center (DLR) gave a class about Object-based image analysis (OBIA) using the eCognition Software for the EAGLE students. The course gives an insight into the theoretical basis of OBIA and using different datasets and tasks...

PhD defense by Thilo Erbertseder

PhD defense by Thilo Erbertseder

Thilo Erbertseder will defend his PhD thesis "Satellite-based analysis of NO2 air pollution: from global to urban aspect" on Wednesday 23rd of July at 3pm in John-Skilton Str. 4a, seminar room 1. All interested staff, students, family and friends are cordially invited...

upcoming PhD defense by Adomas Liepa

upcoming PhD defense by Adomas Liepa

Our PhD student Adomas Liepa will defend his Phd "Potential of Satellite Earth Observation in seasonal monitoring of complex agricultural environments of East Africa" on Thursday 24th of July at 11am. The defense will take place at John Skilton Str. 4a, seminarroom 1....

EORC Staff and EAGLE Students at ESA Living Planet Symposium 2025

EORC Staff and EAGLE Students at ESA Living Planet Symposium 2025

This week, our EORC team and EAGLE MSc students are joining the global Earth observation community at the ESA Living Planet Symposium (LPS) 2025 — one of the most important gatherings for Earth system scientists, remote sensing experts, and space agencies worldwide....