sitsfeats - Satellite Image Time Series Features

R package for extracting metrics from time series of satellite images

The goal of the sitsfeats package is to extract temporal metrics from satellite images time series. The metrics implemented in this version are:

• basics: statistics metrics
• polar: polar metrics

The implementations of this package are based on the stmetrics library in Python.

Installation

You can install the CRAN version with:

install.packages("sitsfeats")

Or the development version of sitsfeats from github with:

library(devtools)
devtools::install_github("oldlipe/sitsfeats")

How to use?

Polar metrics example

# loading sitsfeats and tibble packages
library(sitsfeats)
library(tibble)

# loading example time series
data("timeseries")

polar_metrics <- tibble::tibble(area_q1 = area_q1(timeseries),
                                area_q2 = area_q2(timeseries),
                                area_q3 = area_q3(timeseries),
                                area_q4 = area_q3(timeseries))

polar_metrics
#> # A tibble: 5 × 4
#>   area_q1[,1] area_q2[,1] area_q3[,1] area_q4[,1]
#>         <dbl>       <dbl>       <dbl>       <dbl>
#> 1   29443841.   14207184.   30180681.   30180681.
#> 2   28600327.   11275172.   32912031.   32912031.
#> 3   39626598.   26313817.   39566583.   39566583.
#> 4   37874557.   16023772.   34952318.   34952318.
#> 5   23932473.   15016832.   27336259.   27336259.

Polar metrics visualization

# get one instance from timeseries
one_instance <- timeseries[1,]

# plot first time series example
polar_plot(one_instance)

Getting help

Feel free to create issues if you have any questions.