Gaussian smoothing of raster
Applies a Gaussian smoothing kernel to smooth raster.
raster.gaussian.smooth(x, sigma = 2, n = 5, type = mean, ...)
x |
raster object |
sigma |
standard deviation (sigma) of kernel (default is 2) |
n |
Size of the focal matrix, single value (default is 5 for 5x5 window) |
type |
The statistic to use in the smoothing operator (suggest mean or sd) |
... |
Additional arguments passed to raster::focal |
raster class object of the local distributional moment
This is a simple wrapper for the focal function, returning local statistical moments
Jeffrey S. Evans <jeffrey_evans@tnc.org>
library(raster)
r <- raster(nrows=500, ncols=500, xmn=571823, xmx=616763,
ymn=4423540, ymx=4453690)
proj4string(r) <- crs("+proj=utm +zone=12 +datum=NAD83 +units=m +no_defs")
r[] <- runif(ncell(r), 1000, 2500)
r <- focal(r, focalWeight(r, 150, "Gauss") )
# Calculate Gaussian smoothing with sigma(s) = 1-4
g1 <- raster.gaussian.smooth(r, sigma=1, nc=11)
g2 <- raster.gaussian.smooth(r, sigma=2, nc=11)
g3 <- raster.gaussian.smooth(r, sigma=3, nc=11)
g4 <- raster.gaussian.smooth(r, sigma=4, nc=11)
opar <- par(no.readonly=TRUE)
par(mfrow=c(2,2))
plot(g1, main="Gaussian smoothing sigma = 1")
plot(g2, main="Gaussian smoothing sigma = 2")
plot(g3, main="Gaussian smoothing sigma = 3")
plot(g4, main="Gaussian smoothing sigma = 4")
par(opar)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.