disparityMap: Calculated disparity map between two SAR records

Description Usage Arguments Value See Also Examples

View source: R/disparityMap.r

Description

Calculates a disparity map between two SAR records using window based zero normalized cross correlation (ZNCC).

Usage

1
2
3
disparityMap(master, slave, window.size = 3, search.area.size = 7,
  search.area.shift = c(0, 0), resample.slave = T, window.moving.step = 1,
  run.parallel = F, cores = 4, log = F, log.file = "dispMapLog.txt")

Arguments

master

Object of the SAR-class or a subclass (e.g. Sentinel-class or TSX-class).

slave

Object of the SAR-class or a subclass (e.g. Sentinel-class or TSX-class).

window.size

Integer or vector of integers. Edge Length of quadratic window(s) to calculate zncc. Correlation coefficients are multiplied if a vector is provided.

search.area.size

Integer. Edge Length of quadratic search area in slave image.

search.area.shift

Numeric vector. Pixels in x and y direction that the search area should be shifted. That is how a priori knowledge about disparity can be regarded.

resample.slave

Logical. Should the slave image be resampled to the aggregated master image?

window.moving.step

Integer. Step size to move windows in slave image. Default is 1.

run.parallel

Logical. Run algorithm on more than one cores?

cores

Integer. How many cores should be allocated?

log

Logical. Log output to text file?

log.file

Character. Log file address and name.

Value

Four dimensional array. The first and second dimension give rows and columns, respectively. The third and forth dimension gives disparities in x and y direction. The disparity is measured in pixel.

See Also

zncc

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
data(kili)

master <- aggregate(master, 5)

disp.map <- disparityMap(master, slave, window.size=11, search.area.size=25)

disp.map.lon <- raster(disp.map[,,1])
extent(disp.map.lon) <- extent(master)
plot(disp.map.lon)

disp.map.lat <- raster(disp.map[,,2])
extent(disp.map.lat) <- extent(master)
plot(disp.map.lat)

disp.map.diagonal <- disp.map.lon
values(disp.map.diagonal) <- sqrt(disp.map.lat[]^2 + disp.map.lon[]^2)
plot(disp.map.diagonal)

# to run parallel register cores first, e.g.:
# library(doMC)
# registerDoMC(4)

ragram documentation built on May 2, 2019, 4:42 p.m.