raster_local_max {lidaRtRee}

Statistics of raster local maximum

Description

identifies global maximum and second global maximum from raster (e.g. output from rasters_moving_cor), and computes related statistics. Local maxima can be excluded based on a minimum distance dm to nearest local maximum.

Usage

raster_local_max(r, dm = 2, med1 = 1, med2 = 2, quanta = 0.75, quantb = 0.5)

Arguments

r

SpatRaster. typically output of rasters_moving_cor

dm

numeric. minimum distance between two local maxima in meters

med1

numeric. window radius to compute median value around the maximum position (default: 1m)

med2

numeric. window radius #2 to compute median value around the maximum position (default: 2m)

quanta

numeric. quantile value to compute for raster values (default: 3rd quartile)

quantb

numeric. quantile #2 value to compute for raster values (default: median)

Value

A data.frame with value of maximum, position of maximum, position of second maximum, ratio of max value to 2nd max, ratio of max value to median of neighborhood (size1 and size 2), ratio of max value to raster quantiles 1 and 2

See Also

rasters_moving_cor, coregistration for application to the coregistration of tree inventory data with canopy height models

Examples

# create raster
r_b <- terra::rast(xmin = 0, xmax = 40, ymin =0 , ymax = 40,
resolution = 1, crs = NA)
xy <- terra::xyFromCell(r_b, 1:(nrow(r_b) * ncol(r_b)))

# add Gaussian surfaces
z1 <- 1.5 * exp(-((xy[, 1] - 22)^2 + (xy[, 2] - 22)^2 / 2) / 5)
z2 <- exp(-((xy[, 1] - 20)^2 + (xy[, 2] - 22)^2 / 2) / 3)
z3 <- 1.5 * exp(-((xy[, 1] - 17)^2 + (xy[, 2] - 17)^2 / 2) / 5)
r_b <- terra::rast(cbind(xy, z1 + z2 + z3), type = "xyz")

# create small raster
r_s <- terra::crop(r_b, terra::ext(c(15, 25, 15, 25)))
# offset raster by (-2, -2)
terra::ext(r_s) <- c(13, 23, 13, 23)

rr <- rasters_moving_cor(r_b, r_s, buffer = 6, step = 1)
loc_max <- raster_local_max(rr)
loc_max

# plot raster
terra::plot(rr)
# add location of two local maxima
graphics::points(loc_max[1, c("dx1", "dx2")], loc_max[1, c("dy1", "dy2")],
  cex = c(1, 0.5), pch = 3
)