iMad: iMad (optimized port)

In imad: Iteratively Re-weighted Multivariate Alteration Detection (iMAD)

Description

Perform iteratively re-weighted multivariate alteration detection.

Usage

  iMad(inDataSet1, inDataSet2, pos, mask1, mask2,
    mask1_band = 1, mask2_band = 1, output_basename,
    format = "raster", maxiter = 100, lam = 0,
    delta = 0.001, corr_thresh = 0.001,
    auto_extract_overlap = TRUE,
    reuse_existing_raster = TRUE, force_extent = TRUE,
    enable_snow = FALSE, cl = NULL, verbose = FALSE,
    timing = FALSE, inmemory = FALSE, debug = FALSE,
    debug_outputs = NULL, ...)

Arguments

inDataSet1	A Raster* object of the first image.
inDataSet2	A Raster* object of the second image.
pos	Integer vector. A vector of bands to use from each image. Default is use all bands.
mask1	(Optional) A Raster* object representing a mask to be used for inDataSet1 or a numeric value to be used as the mask value.
mask2	(Optional) A Raster* object representing a mask to be used for inDataSet2 or a numeric value to be used as the mask value.
mask1_band	(Optional) The band from inDataSet1 to use for masking (only if class(mask1)=="numeric").
mask2_band	(Optional) The band from inDataSet2 to use for masking (only if class(mask1)=="numeric").
output_basename	Character. The basename (including path) for the output files.
format	Character. The output format of the rasters (see ?writeFormats). Default is "raster".
maxiter	Numeric (>= 1). The maximum number of iterations. Default is 100.
lam	Numeric. The penalization function. CURRENTLY UNSUPPORTED.
corr_thresh	Numeric. Used for situations where the canonical correlates are all nearly 1.0 (how close to 1.0 does it need to be to stop).
delta	Numeric. The smallest change in canonical correlates to end the program.
auto_extract_overlap	Logical. Extract the overlap zones between the images?
reuse_existing_raster	Logical. If the algorithm detects pre-create overlaps, use them?
force_extent	Logical. Attempt to force the input files (and masks) to be the same extent?
enable_snow	Logical. Use clusterR to (potentially) speed up calculations on a cluster? Default=FALSE. EXPERIMENTAL.
cl	Cluster. If not assigned, the program will attempt to figure it out. Use beginCluster() to create a cluster.
verbose	Logical. Print out debugging information?
...	Passed to various raster functions (see writeRaster). Important ones include format= and overwrite=TRUE/FALSE. datatype should be left as 'FLT4S' for proper functioning.

Value

Returns a RasterStack object where the first layer is the chisquare image, and the subsequent layers are the iMad layers.

Author(s)

Mort Canty (original code) and Jonathan A. Greenberg (R port).

References

• Canty, M.J. and A.A. Nielsen. 2008. Automatic radiometric normalization of multitemporal satellite imagery with the iteratively re-weighted MAD transformation. Remote Sensing of Environment 112:1025-1036.

• Nielsen, A.A. 2007. The regularized iteratively reweighted MAD method for change detection in multi- and hyperspectral data. IEEE Transactions on Image Processing 16(2):463-478.

See Also

writeRaster

imad documentation built on May 2, 2019, 6:05 p.m.