Warp tool

Instructs Bentley I/RAS B to perform a Helmert, Projective, or Affine warp operation on the active layer or all layers. When you select Modify, then click Warp on the Edit pulldown menu (or open the Warp icon on the Modify palette), the I/RAS B Warp dialog box appears.

A warp is a two-dimensional topological transformation in which a source area is fitted to a destination area. Usually the Warp feature is used to counteract some influence on the original raster data. For example, a satellite photograph that includes asymmetrical curvature may need to be fitted to a square grid for mapping purposes. Any data captured with a camera may also have lens distortion errors that add another dimension to error correction.

In effect, Warp does reverse interpolation to remove unwanted distortions and arrive at a desired raster data form. The source is the raw, distorted data, while the destination is the desired data after the warp operation.

The relationship between source and destination is specified by the model of the warp (Helmert, Projective, or Affine) and by a series of source/destination point pairs. These pairs specify the relationship between the original drawing and the drawing after warping.

To perform a warp, first determine the nature of the data distortion. Is the data bent, or just stretched? How many bends does the data exhibit? Determine the model and order of warp to use. It is best to use the simplest and lowest order warp that will achieve the desired effect. Using high order warps can generate unpredictable results in areas away from the source/destination pairs. If a Helmert warp will suffice to correct a particular drawing, use Helmert warping instead of Affine or Projective warping. Projective warping is more time-consuming than a four-point first order Affine warp.

Enter as many source/destination pairs as it takes to accomplish the warp. Think of destination data points as tent stakes that pull on the fabric of the raster data. Entering points close together has the effect of weighting an area. It is often important to include one or more pairs of points near the center of the raster data, to better anchor the data while you “pull on the edges.”

The warp transformation performs error optimization using a least square fit method to determine the effect of each pair of data points. When more than the minimum number of source/destination points is entered, it is often not possible to fit each source point precisely to its destination point. In this case a residual value predicts how much error each destination point will have. The X, Y, and total error distances are displayed in the Warp dialog box before executing the warp operation. Depending on your purpose for warping, you may wish to delete and reenter source/destination pairs in order to affect the distribution of residual errors. After you are satisfied with residual distribution, select Perform Warp to start the raster operation.

• Warp Speed - Helmert and first order Affine are faster than other warp methods. This is because Helmert and first order Affine warps use a highly optimized linear resampling algorithm, while other warp methods require more complex non-linear calculations.

• Helmert Warp - A Helmert warp operation rotates and scales the raster data. A minimum of two pairs of source/destination points are required to specify a helmert warp.

• Projective Warp - Projective warping is designed to correct the effects of roll, pitch, and/or yaw in satellite and/or aerial photographs. A minimum of four source/destination point pairs are required. Projective warping is more time consuming than a similar four-point first order Affine warp or Helmert warp.

• Affine Warp - Affine warping stretches data in one or more directions, depending on the warp order. The Order parameter determines the nature of the transformation by setting the mathematical model for the warp. First order polynomials are simple linear stretching (commonly referred to as rubber sheeting). Second order Affine warps use second order polynomials, third order warps use third order polynomials, and so on. A second order polynomial has one bend, such as a parabola. A third order polynomial has an S-shape, and so forth. The order of the Affine warp determines the minimum number of source/destination point pairs.

  Order	Minimum number of source/destination point pairs
  1	13
  2	6
  3	10
  4	15
  5	21

  The Undo feature is available for the Rectangle and Polygon area options. If Undo is disabled under the Settings/Undo command, then Warp/Undo is ignored.

• Placement Effects - When the entire raster drawing is selected for Helmert and first order Affine transformations, the placement of the drawing is changed by the warp operation. When the Rectangle or Polygon options are selected, the placement of the drawing is not affected by the warp operation.

Tool Settings	Effect
Transformation Model	Select model for the warp operation Helmert — Selects Helmert warping. Projective — Selects Projective warping. Affine 1 — Selects first order Affine warping. Affine 2 — Selects second order Affine warping. Affine 3 — Selects third order Affine warping. Affine 4 — Selects fourth order Affine warping. Affine 5 — Selects fifth order Affine warping. The control points used to control the building of a model are displayed below in the Mininum Points and Number of Points sections.
Warp Area	Choose area for warp operation: Drawing — Entire raster drawing is selected for warp operation. Rectangle — Rectangular area destination selection. Polygon — Polygonal area destination selection.
Layers	Choose active layer(s): Active — Warp data in active layer only. All — Warp data in all visible layers.
Undo	Select Undo On or Undo Off: On — Enables the Undo feature for the Warp command (not available for warping of entire drawing). Off — Disables the Undo feature for the Warp command.
Smooth	Select Smooth On or Smooth Off: On — Enables smoothing of data before and after warping. Off — Disables smoothing.
Collect Points	Use to select the point for the Warp operation.
Delete Point	Select Delete Point if you want to remove a point pair. Enter the index of the point pair to remove.
Perform Warp	Starts the warp operation.
Cancel	Exit the I/RAS B Warp dialog.

To use Warp