m.gcp.filter - GRASS GIS manual

DESCRIPTION
NOTES
SEE ALSO
AUTHOR

NAME

m.gcp.filter - Filter Ground Control Points (GCPs).

KEYWORDS

miscellaneous, transformation, GCP

SYNOPSIS

m.gcp.filter

m.gcp.filter --help

m.gcp.filter [-bdu] group=name order=integer threshold=float [iterations=integer] [--help] [--verbose] [--quiet] [--ui]

Flags:

-b: Use backward transformations (default: forward transformations)
-d: Use GCP deviation (default: RMS)
-u: Update GCPs
--help: Print usage summary
--verbose: Verbose module output
--quiet: Quiet module output
--ui: Force launching GUI dialog

Parameters:

group=name [required]: Name of input imagery group
order=integer [required]: Rectification polynomial order; Options: 1-3
threshold=float [required]: Filtering threshold in CRS units
iterations=integer: Maximum number of iterations

DESCRIPTION
NOTES
SEE ALSO
AUTHOR

DESCRIPTION

m.gcp.filter filters GCPs using the distance between the computed coordinates and the actual coordinates of each GCP. The result of the filtering (number of points to use after filtering, number of points filtered out, final RMS error) is printed out.

The transformation equations are computed anew for each iteration. The GCP with the largest distance between the computed coordinates and the actual coordinates is deactivated and the next iteration is started. Filtering stops when the overall RMS error is below the given threshold or when the number of active GCPs is equal to the required minimum number of points or when the optional maximum number of iterations (option iterations) has been reached. With the -d flag, the largest GCP error is used instead of the overall RMS error.

m.gcp.filter uses by default the results of forward transformations (source to target) for filtering. With the -b flag, the results of backward transformations (target to source) are used.

The status of GCPs (active / inactive) is only updated with the -u flag. GCPs that have been filtered out will be deactivated, not deleted.

NOTES

The transformations are:

order=1:

    e = [E0 E1][1].[1]
        [E2  0][e] [n]

    n = [N0 N1][1].[1]
        [N2  0][e] [n]

order=2:

    e = [E0 E1 E3][1 ] [1 ]
        [E2 E4  0][e ].[n ]
        [E5  0  0][e²] [n²]

    n = [N0 N1 N3][1 ] [1 ]
        [N2 N4  0][e ].[n ]
        [N5  0  0][e²] [n²]

order=3:

    e = [E0 E1 E3 E6][1 ] [1 ]
        [E2 E4 E7  0][e ].[n ]
        [E5 E8  0  0][e²] [n²]
        [E9  0  0  0][e³] [n³]

    n = [N0 N1 N3 N6][1 ] [1 ]
        [N2 N4 N7  0][e ].[n ]
        [N5 N8  0  0][e²] [n²]
        [N9  0  0  0][e³] [n³]

["." = dot-product, (AE).N = N'EA.]

In other words, order=1 and order=2 are equivalent to order=3 with the higher coefficients equal to zero.

AUTHOR

Markus Metz

SOURCE CODE

Available at: m.gcp.filter source code (history)

Latest change: Thursday Mar 20 21:36:57 2025 in commit: 7286ecf7af235bfd089fb9b1b82fb383cf95f3fc