The Transform feature reprojects an existing georeferenced image from its original coordinate system to another one. The resulting transformed or destination image and associated reference file can be used in many GIS, CAD and desktop mapping systems including MAPublisher for Adobe Illustrator.
Transform Dialog Box
When an image has a coordinate system specified, the Transform button in the Geographic Imager panel will be enabled. Click the Transform button to open the Transform dialog box—which displays the current geographic parameters of an image and settings to transform it.
Transform button
Destination Coordinate System
Destination
To set a destination coordinate system, click the coordinate system hyperlink. The Specify Destination Coordinate System dialog box opens and displays coordinate systems and projections available in the Geographic Imager data source. This dialog box is similar to the Specify Source Coordinate System dialog box, however, here you are specifying a destination coordinate system instead of a source. The destination coordinate system is the coordinate system you want to transform the image to. Explore the categories to find and select the desired destination coordinate system.
Coordinate systems are tested upon selection/transformation to verify their suitability for a referenced image. A warning will appear if the coordinate system appears to be partially or fully incompatible (based on the coordinate system envelope).
To set the coordinate system to be the same as that of another currently open and referenced image, choose the Same As option and the desired image from the drop-down list. If there are no other referenced images open this option will be disabled. Modifications to pixel size, resampling method, and strip width are available as options (explained in detail below).
Alternatively, click the From File button to use a coordinate system definition from a coordinate system definition file (.wkt or .prj).
Datum shift
Geographic Imager automatically selects an appropriate datum shift for the transformation if required. If the selected datum shift is not the desired one, click the coordinate system hyperlink to open the Select Datum Shift dialog box. Datum shift types are displayed in the left category column titled Matching Datum Transformations.
Each category contains the associated datum shift(s) for that reference network. To view the information associated with the datum shift(s), click the View Datum Transformation button. If a second datum shift is necessary, reference networks will be listed in a subcategory called Using (coordinate system) as Intermediate Datum. The View Second Datum Transformation button will be enabled to view the secondary datum shift information.
Due to the complex nature of performing such transformations, this process may require extended processing time depending on the original image size, destination pixel size, strip width, resampling method, and other parameters. For multiple transformations, consider running scripts or actions to automate tasks.
Multi-layer transformations can be transformed without flattening the image, maintaining the original layers. Existing transparency is properly maintained during transformation. To guarantee a seamless destination image (i.e. with no gaps between the adjacent layers), make sure there is a sufficient pixel overlap between the layers (e.g. when tiling the original image). Using the Nearest-Neighbor method or a smaller strip width may also solve the problem.
Multi-spectral images are supported and maintained for transformations, including support for infrared (RGBI), alpha, and spot channels.
Layer masks are ignored during the Transformation process. Layers masks are retained, however, they may cause unforeseen errors if used in a workflow. It is recommended that layers masks be removed before any transformations are performed.
When saving an image after a transformation has been performed, Geographic Imager will update and overwrite the original reference file. In order to keep the original image and reference files, use the Save As command with a new file name or file location rather than using the Save command.
Pixel options
The Pixel Size indicates the pixel resolution as the ground distance covered by each pixel of the image. The pixel dimension and unit (corresponding to the selected destination coordinate system) are displayed. Pixels are commonly square for projected coordinate systems. In a geodetic coordinate system, pixels are not necessarily square due to varying sizes of degrees of latitude depending on geographic location. Therefore, the x-to-y ratio of a pixel could be determined by the value of the Central Latitude which is commonly estimated by the center of the image.
Modify Pixel Dimensions
Click the Modify button in the Transformation dialog box to open the Modify Pixel Dimensions dialog box.
The following are the available methods to calculate pixel dimensions:
|
Preserve existing |
Keeps the pixel dimension ratio the same as in the source image. It keeps the transformed geographic extents close to the original. |
|
By center of output image |
Calculate the pixel ratio based on the central latitude of the outgoing image which will be the latitude of the true scale. Choose this option to minimize the distance of distortion. |
|
Keep pixels square |
Ensures the pixels of the transformed image are square. |
|
User-specified |
Specify a custom x-to-y ratio. Enables Pixel Size text boxes above. |
Certain methods such as Preserve existing and By the center of output, image methods are not available for all transformations. To use the Preserve existing method, the transformation needs to be either Geodetic to Geodetic or Projected to Projected. If the destination image is geodetic, “y” pixel dimension is effectively determined by the choice of the central latitude (which is also the latitude of the true scale), therefore the option to set the pixel ratio By center of output image is only available when the transformation outcome is indeed Geodetic.
To adjust the pixel precision (number of decimals shown), go to General Preferences.
Advanced Options
Resampling Methods
There are several options for resampling the image during transformation:
|
Bicubic (smooth gradients) |
Takes a weighted average value of the sixteen pixels closest to the transformed location in the source image. |
|
Bilinear |
Takes a weighted average value of the four pixels closest to the transformed location in the source image. This method results in a smoother image than the nearest neighbor method but at the expense of more processing time. |
|
Nearest Neighbor (hard edge) |
Takes the value of the pixel that is closest to the transformed location in the source image. This is the fastest method in terms of processing time and is the method to use to preserve a colour panel during image transformation. It is also the best method to preserve original colours in some non-RGB colour modes, such as CMYK, by avoiding the internal conversion to/from RGB. |
|
Automatic |
Adobe Photoshop chooses the resampling method based on the document type and whether the document is scaling up or down. |
|
Preserve Details (enlargement) |
A method for enlarging images that attempts to smooth out noise. |
|
Bicubic Smoother (enlargement) |
A method for enlarging images based on Bicubic interpolation but designed to produce smoother results. |
|
Bicubic Sharper (reduction) |
A method for reducing the size of an image based on Bicubic interpolation with enhanced sharpening. This method maintains the detail in a resampled image. If Bicubic Sharper oversharpens some areas of an image, try using Bicubic. |
A coordinate system transformation will resample an image, therefore any image analysis or classification should be performed prior to a transformation.
Layers
There are several options to handle layers during transformation:
|
Leave intact |
Maintains the original layer structure. |
|
Merge raster layers |
Combines only raster layers while maintaining vector layers, for example text and transparency. Choose this option to guarantee seamless edges between different layers, especially if they don’t have overlapping margins.
|
|
Merge layers |
Combines all layers into one layer, maintaining transparency.
|
Trim transparent edges
Enabled by default, this option trims any excess transparency while maintaining the extents of the image.
Precision and Strip width
To provide necessary transformation performance, Geographic Imager transforms one horizontal strip of the image at a time. Change the Precision to adjust the strip width being transformed:
|
Maximum (slow; best for World Projections) |
Recommend when performing transformations on images with the extent of the world. |
|
High |
Uses a strip width of 10 pixels—high quality results and slightly slower transformation performance. |
|
Normal |
The default setting using a strip width of 50 pixels—a reasonable value for the majority of raster imagery combining good quality results and high transformation performance. |
|
Medium (fast) |
Uses a strip width of 250 pixels—good quality results and faster transformation performance. |
|
Custom |
Set a custom strip width. |
Some images may be optimized by setting a custom strip width, which may be especially true when working with very small/world scale data, high latitude/polar area imagery or projections that introduce severe spatial distortions. If the result contains visible horizontal stripping artifacts or transparent gaps, choose a smaller strip width or the Maximum precision setting. Choosing a smaller strip width will result in a higher-precision transformation, but the transformation process may be slower. Alternatively, for low-deformation, large-scale data or for a quick preview of transformation results, a larger strip width can be used.
For example, if an image being transformed displays a spherical curve, a smaller strip width may be necessary to maintain a smooth image edge. If a larger strip width is used there may a be a stepped edge to the image.
Related topics
Comments
0 comments
Please sign in to leave a comment.