GRASS GIS manual: v.rectify

Note: A new GRASS GIS stable version has been released: GRASS GIS 7.8, available here.
Updated manual page: here

NAME

v.rectify - Rectifies a vector by computing a coordinate transformation for each object in the vector based on the control points.

KEYWORDS

vector, rectify, level1

SYNOPSIS

v.rectify

v.rectify --help

v.rectify [-3orb] input=name output=name [group=name] [points=name] [rmsfile=name] [order=integer] [separator=character] [--overwrite] [--help] [--verbose] [--quiet] [--ui]

Flags:

-3: Perform 3D transformation
-o: Perform orthogonal 3D transformation
-r: Print RMS errors; Print RMS errors and exit without rectifying the input map
-b: Do not build topology; Advantageous when handling a large number of points
--overwrite: Allow output files to overwrite existing files
--help: Print usage summary
--verbose: Verbose module output
--quiet: Quiet module output
--ui: Force launching GUI dialog

Parameters:

input=name [required]: Name of input vector map; Or data source for direct OGR access
output=name [required]: Name for output vector map
group=name: Name of input imagery group
points=name: Name of input file with control points
rmsfile=name: Name of output file with RMS errors (if omitted or '-' output to stdout
order=integer: Rectification polynomial order (1-3); Options: 1-3; Default: 1
separator=character: Field separator for RMS report; Special characters: pipe, comma, space, tab, newline; Default: pipe

DESCRIPTION
- Coordinate transformation and RMSE
SEE ALSO
AUTHOR

DESCRIPTION

v.rectify uses control points to calculate a 2D or 3D transformation matrix based on a first, second, or third order polynomial and then converts x,y(, z) coordinates to standard map coordinates for each object in the vector map. The result is a vector map with a transformed coordinate system (i.e., a different coordinate system than before it was rectified).

The -o flag enforces orthogonal rotation (currently for 3D only) where the axes remain orthogonal to each other, e.g. a cube with right angles remains a cube with right angles after transformation. This is not guaranteed even with affine (1^st order) 3D transformation.

Great care should be taken with the placement of Ground Control Points. For 2D transformation, the control points must not lie on a line, instead 3 of the control points must form a triangle. For 3D transformation, the control points must not lie on a plane, instead 4 of the control points must form a triangular pyramid. It is recommended to investigate RMS errors and deviations of the Ground Control Points prior to transformation.

2D Ground Control Points can be identified in g.gui.gcp.

3D Ground Control Points must be provided in a text file with the points option. The 3D format is equivalent to the format for 2D ground control points with an additional third coordinate:

 x y z east north height status

where x, y, z are source coordinates, east, north, height are target coordinates and status (0 or 1) indicates whether a given point should be used. Numbers must be separated by space and must use a point (.) as decimal separator.

If no group is given, the rectified vector will be written to the current mapset. If a group is given and a target has been set for this group with i.target, the rectified vector will be written to the target location and mapset.

Coordinate transformation and RMSE

The desired order of transformation (1, 2, or 3) is selected with the order option. v.rectify will calculate the RMSE if the -r flag is given and print out statistcs in tabular format. The last row gives a summary with the first column holding the number of active points, followed by average deviations for each dimension and both forward and backward transformation and finally forward and backward overall RMSE.

2D linear affine transformation (1st order transformation)

x' = a1 + b1 * x + c1 * y

y' = a2 + b2 * x + c2 * y

3D linear affine transformation (1st order transformation)

x' = a1 + b1 * x + c1 * y + d1 * z

y' = a2 + b2 * x + c2 * y + d2 * z

z' = a3 + b3 * x + c3 * y + d3 * z

The a,b,c,d coefficients are determined by least squares regression based on the control points entered. This transformation applies scaling, translation and rotation. It is NOT a general purpose rubber-sheeting, nor is it ortho-photo rectification using a DEM, not second order polynomial, etc. It can be used if (1) you have geometrically correct data, and (2) the terrain or camera distortion effect can be ignored.

Polynomial Transformation Matrix (2nd, 3d order transformation)

v.rectify uses a first, second, or third order transformation matrix to calculate the registration coefficients. The minimum number of control points required for a 2D transformation of the selected order (represented by n) is

((n + 1) * (n + 2) / 2)

or 3, 6, and 10 respectively. For a 3D transformation of first, second, or third order, the minimum number of required control points is 4, 10, and 20, respectively. It is strongly recommended that more than the minimum number of points be identified to allow for an overly-determined transformation calculation which will generate the Root Mean Square (RMS) error values for each included point. The polynomial equations are determined using a modified Gaussian elimination method.

AUTHOR

Markus Metz

based on i.rectify

Last changed: $Date$

SOURCE CODE

Available at: v.rectify source code (history)