Scroll to navigation

pksetmask(1) pksetmask(1)

NAME

pksetmask - program to apply mask image (set invalid values) to raster image

SYNOPSIS


pksetmask
-i input -m mask [-msknodata value] -o output [options] [advanced options]

DESCRIPTION

pksetmask sets a mask provided with option -m to an input raster dataset. The default operator is '='. Values in the input raster data where the mask has a nodata value (set with the option -msknodata) will then be set to nodata (set with -nodata). Other operators are less than (--operator '<') and larger than (--operator '<').

OPTIONS

Input image
Mask image(s)
Mask value(s) where image has nodata. Use one value for each mask, or multiple values for a single mask.
Output mask file
nodata value to put in image if not valid
verbose

Advanced options

Operator: < = > !. Use operator for each -msknodata option
Data type for output image ({Byte / Int16 / UInt16 / UInt32 / Int32 / Float32 / Float64 / CInt16 / CInt32 / CFloat32 / CFloat64}). Empty string: inherit type from input image
Output image format (see also gdal_translate(1)).
Creation option for output file. Multiple options can be specified.
colour table in ASCII format having 5 columns: id R G B ALFA (0: transparent, 255: solid)

EXAMPLE

Using a single mask

With a single mask you can provide as many triples (--operator, --msknodata, --nodata) as you wish. All operators work simultaneously on that mask. Caution: the first operator that tests true will be selected. This is explained in the next example:

pksetmask -i input.tif -m mask.tif --operator='>' --msknodata 0 --nodata 0 --operator='>' --msknodata 10 --nodata 10 -o output.tif

Warning: second operator will never test true as first will supersede!

pksetmask -i input.tif -m mask.tif --operator='>' --msknodata 10 --nodata 10 --operator='>' --msknodata 0 --nodata 1 -o output.tif

OK: values above 10 will be 10, values between 0 and 10 will be 1

Using multiple masks

With multiple masks, you can use one triple (--operator, --msknodata, --nodata) for each corresponding mask (following the same order of input). If the number of triples is not equal to the number of masks, then only the first triple is used for all masks simultaneously

pksetmask -i input.tif -m mask1.tif --operator '>' --msknodata 250 --nodata 1 -m mask2.tif --operator '>' --msknodata 100 --nodata 2 -o output.tif

If mask1.tif is above 250, the output will be 1. If mask2 is above 100, the output will be 2. If both operators test true, the first will supersede (output will be 1)

pksetmask -i input.tif -m mask1.tif -m mask2.tif --operator '>' --msknodata 250 --nodata 1 -o output.tif

If either mask1.tif or mask2.tif is above 250, the output will be 1

More examples

pksetmask -i input.tif -m mask.tif -o output.tif -ot Byte --msknodata 0 -nodata 255

copy pixel values from input.tif to output.tif, applying mask.tif, setting all values to 255 where mask is 0.

pksetmask -i input.tif -m mask.tif -o output.tif -ot Byte --msknodata 1 -nodata 255 --operator '!'

copy values from input.tif to output.tif, but set all values to 255 if mask is not 1

pksetmask -i input.tif -m mask1.tif -m mask2.tif -o output.tif -ot Byte --msknodata 0 -nodata 255

Application of two masks. Copy pixel values from input.tif to output.tif, setting all values to 255 where either mask is 0.

pksetmask -i input.tif -m mask.tif -o output.tif -ot Byte --msknodata 0 --msknodata 1 -nodata 255 -nodata 255

copy pixel values from input.tif to output.tif, applying single masks, setting all values to 255 where mask is either 0 or 1.

FAQ

Q1. I want to mask my input image (a byte image with values between 0 and 254) with a mask that only covers a spatial subset of the input image. Within the spatial subset of the primary mask, all pixels must be set to 0 where the primary mask equals 1. Outside the spatial subset I want to set all pixel values to 255.

A1. This can be done using two masks, selecting the input image as the secondary mask. Choose the secondary operator acting on the secondary mask such that the condition is always true (e.g, < 255).

06 December 2020