Syntax

   FACTORY_DEF <ReaderKeyword> RasterSharpenerFactory
   [FACTORY_NAME <factory name>]
   [INPUT FEATURE_TYPE <factory type>
      [<attribute name> <attribute value>]*
      [<feature function>]*]*
   [SHARPENING_FILTER_MODE (Fine|Coarse)]
   [SHARPENING_FACTOR <sharpening factor>]
   [SHARPENING_ITERATIONS <number of iterations>]
   [PRESERVE_BAND_INTERPRETATION (Yes/No)]
   [OUTPUT SEGMENTED FEATURE_TYPE <feature type>
      [<attribute name> <attribute value>]*
      [<feature function>]*]*

Overview

   The RasterSharpener accepts features with raster geometry. All other geometry types will be
   rejected.
   

   Each input raster feature will be processed independently.
   

   The RasterSharpener enhances the features of the and image (borders lines and curves), while
   it reduces noise.

   It's useful as pre processing step for the RasterSegmenter.

Bands, cell type and Palettes

   Rasters may have one band or multiple bands. The RasterSharpener can handle up to 4 bands.

   All raster cell types are accepted as input but the algorithm internally uses Real64 data.
   A preserve interpretation option is provided, that, when enabled, converts the raster back 
   to the same interpretation as the input raster.
   

   Rasters may have no palette, one palette, or multiple palettes. For more information on raster
   structure, see Rasters (IFMERaster). Input rasters with string palette values on cannot be 
   resolved and will be rejected.
   

   The RasterSharpener will resolve palettes on input rasters prior to
   processing.


 Options

      SHARPENING_FILTER_MODE
         Coarse:
            Perona-Malik diffusion.

            Eliminate much of the noise and minor features of the image while sharpening heavily 
            stronger borders.

            Sensitive to contrast

         Fine(Default):
            A variant of Perona-Malik diffusion.

            Preserve much of the details of the image while removing noise in the flatter areas.
      
            Less sensitive to contrast.

            This mode takes roughly twice as much time per iteration as the Coarse option.

         Required.

      SHARPENING_FACTOR

         This floating point parameter determines the sensitivity of the conductance equation in the
         Anisotropic Diffusion Parameter.

         Generally, lower values will cause better preservation of image characteristics (such as
         sharp changes and borders).
         
         In Fine mode, useful values tend to be close to 3.0. In coarse mode typical values range from 0.5
         to 2.0. The values values depend heavily on the desired result and number of iterations the filter
         is run for.
         

         Required.

      

      SHARPENING_ITERATIONS

         How many times the Anisotropic Diffusion Filter is going to be run.        

         More iterations will sharpen more the features of the image that are preserved by the
         filter. 

         This parameter accepts a positive integer. Typical values range between 1 and 15.
         Usually the Fine option tends to be more useful with values closer to the lower
         end (such as 5), while with the Coarse mode values closer to 15 can be more useful

         Default is 10.

         Required.

      PRESERVE_BAND_INTERPRETATION
         
         No (default): 
            The bands of the sharpened raster will have data type Real64, which is the
            datatype internally used for computation. This option can be faster and is recommended
            if the output is being used by factories that also process internally in Real64, such as
            the RasterSegmenter.

         Yes:
            After computation, the bands will be converted back to their original interpretation.
            This option can be slower but it's useful when the input raster already as the desired
            interpretation.

         Required.


Output Tags

   The RasterSharpenerFactory supports the following output tags.

   SHARPENED

      Output the segmented raster.

   REJECTED

      If there are any incorrect transformer parameters or the input is incorrect, the feature with
      that raster will be rejected