.

GDAL conversion

gdal2xyz

Kuvaus

<put algorithm description here>

Parameters

Input layer [raster]
<put parameter description here>
Band number [number]

<put parameter description here>

Default: 1

Tulokset

Output file [table]
<put output description here>

Console usage

processing.runalg('gdalogr:gdal2xyz', input, band, output)

See also

PCT to RGB

Kuvaus

Convert an 8bit paletted image to 24bit RGB

This utility will convert a pseudocolor band on the input file into an output RGB file of the desired format.

Parameters

Input layer [raster]
Input 8bit raster image
Band to convert [selection]

Band to convert to RGB

Options:

  • 0 — 1
  • 1 — 2
  • 2 — 3
  • 3 — 4
  • 4 — 5
  • 5 — 6
  • 6 — 7
  • 7 — 8
  • 8 — 9
  • 9 — 10
  • 10 — 11
  • 11 — 12
  • 12 — 13
  • 13 — 14
  • 14 — 15
  • 15 — 16
  • 16 — 17
  • 17 — 18
  • 18 — 19
  • 19 — 20
  • 20 — 21
  • 21 — 22
  • 22 — 23
  • 23 — 24
  • 24 — 25

Default: 0

Tulokset

Output layer [raster]
24bit RGB raster image

Console usage

processing.runalg('gdalogr:pcttorgb', input, nband, output)

Polygonize (raster to vector)

Kuvaus

Creates vector polygons for all connected regions of pixels in the raster sharing a common pixel value. Each polygon is created with an attribute indicating the pixel value of that polygon. The algorithm is derived from the GDAL polygonize utility .

Parameters

Input layer [raster]
Input raster layer.
Output field name [string]

Defines the field name in which the attributes of the connected regions are entered. If no field name is defined the name will be ‘DN’.

Default: DN

Tulokset

Output layer [vector]
Output vector layer. Default format is *.shp .

Console usage

processing.runalg('gdalogr:polygonize', input, field, output)

See also

Rasterize (vector to raster)

Kuvaus

Converts vector geometries (points, lines and polygons) into a raster image. The algorithm is derived from the GDAL rasterize utility .

Parameters

Input layer [vector: any]
Input vector layer with point, line or polygon geometries.
Attribute field [tablefield: any]
Defines the attribute field from which the attributes for the pixels should be chosen.
Write values inside an existing raster layer(*) [boolean]

If activated writes the results into an existing raster layer.

Default: False

Set output raster size (ignored if above option is checked) [selection]

If the above option is not checked either the output size is in pixels (0) or the output resolution is in map units per pixel (1).

Options:

  • 0 — Output size in pixels
  • 1 — Output resolution in map units per pixel

Default: 1

Horizontal [number]

Sets the horizontal resolution of the output raster.

Default: 100.0

Vertical [number]

Sets the vertical resolution of the output raster.

Default: 100.0

Raster type [selection]

Defines the type of the resulting raster image.

Options:

  • 0 — Byte
  • 1 — Int16
  • 2 — UInt16
  • 3 — UInt32
  • 4 — Int32
  • 5 — Float32
  • 6 — Float64
  • 7 — CInt16
  • 8 — CInt32
  • 9 — CFloat32
  • 10 — CFloat64

Default: 0

Tulokset

Output layer: mandatory to choose an existing raster layer if the (*) option is selected [raster]
Output raster layer. Give in here the name of the raster layer in which the results should be written (if this option is used).

Console usage

processing.runalg('gdalogr:rasterize', input, field, writeover, dimensions, width, height, rtype, output)

See also

Muunna RGB PCT:ksi

Kuvaus

Converts a 24bit RGB image into a 8bit paletted. Computes an optimal pseudo-color table for the given RGB-image using a median cut algorithm on a downsampled RGB histogram. Then it converts the image into a pseudo-colored image using the color table. This conversion utilizes Floyd-Steinberg dithering (error diffusion) to maximize output image visual quality. If you want to classify a raster map and want to reduce the number of classes it can be helpful to downsample your image with this algorithm before. The algorithm is derived from the GDAL rgb2pct utility .

Parameters

Input layer [raster]
Input RGB image raster.
Number of colors [number]

The number of colors the resulting image will contain. A value from 2-256 is possible.

Default: 2

Tulokset

Output layer [raster]
8-bit output raster.

Console usage

processing.runalg('gdalogr:rgbtopct', input, ncolors, output)

See also

Translate (convert format)

Kuvaus

<put algorithm description here>

Parameters

Input layer [raster]
<put parameter description here>
Set the size of the output file (In pixels or %) [number]

<put parameter description here>

Default: 100

Output size is a percentage of input size [boolean]

<put parameter description here>

Default: True

Nodata value, leave as none to take the nodata value from input [string]

<put parameter description here>

Default: none

Expand [selection]

<put parameter description here>

Options:

  • 0 — none
  • 1 — gray
  • 2 — rgb
  • 3 — rgba

Default: 0

Output projection for output file [leave blank to use input projection] [crs]

<put parameter description here>

Default: None

Subset based on georeferenced coordinates [extent]

<put parameter description here>

Default: 0,1,0,1

Copy all subdatasets of this file to individual output files [boolean]

<put parameter description here>

Default: False

Additional creation parameters [string]

Optional.

<put parameter description here>

Default: (not set)

Output raster type [selection]

<put parameter description here>

Options:

  • 0 — Byte
  • 1 — Int16
  • 2 — UInt16
  • 3 — UInt32
  • 4 — Int32
  • 5 — Float32
  • 6 — Float64
  • 7 — CInt16
  • 8 — CInt32
  • 9 — CFloat32
  • 10 — CFloat64

Default: 5

Tulokset

Output layer [raster]
<put output description here>

Console usage

processing.runalg('gdalogr:translate', input, outsize, outsize_perc, no_data, expand, srs, projwin, sds, extra, rtype, output)

See also