Skip to contents

Satellite measurements are divided into indivisible units called tiles. The mosaic function binds and crops the tiles to generate a single image of the region of interest for each date.

Usage

rsat_mosaic(x, ...)

# S4 method for rtoi
rsat_mosaic(x, ...)

# S4 method for records
rsat_mosaic(
  x,
  out_path,
  db_path,
  bfilter = c(),
  warp = "extent",
  region,
  overwrite = FALSE,
  verbose = FALSE,
  ...
)

Arguments

x

an rtoi object.

...

additional arguments.

out_path

path to save the mosaicked images. By default, the path is defined by x.

db_path

path to the database. By default, the path is defined by x.

bfilter

a vector of bands to. If not supplied, all are used.

warp

character. If equal to "extent", it also crops the images around the rtoi. Use "" otherwise.

region

an sf object. Region for cropping the images around. By default, the path is defined by x.

overwrite

logical argument. If TRUE, overwrites the existing images with the same name.

verbose

boolean. If TRUE show verbose output. Default FALSE

Value

nothing. Mosaics the downloaded images and stored them on the hard disk

Examples

if (FALSE) {
library(rsat)

# load navarre sf from the package
data(ex.navarre)

# set the credentials
set_credentials("username", "password")

# path where the region is stored
rtoi.path <- tempdir()
# path where downloads are stored
db.path <- file.path(tempdir(), "DATABASE")
navarre <- new_rtoi(
  "Navarre",
  ex.navarre,
  rtoi.path,
  db.path
) #'
# Landsat-5
rsat_search(
  region = navarre,
  product = "LANDSAT_TM_C1",
  dates = as.Date("1988-08-01") + seq(1, 35)
)
rsat_download(navarre)

rsat_mosaic(navarre, overwrite = T)

rsat_list_data(navarre)
}