org.deegree.tools.rendering.dem.filtering.filters
Class SmoothingFilter

java.lang.Object
  extended by org.deegree.tools.rendering.dem.filtering.filters.SmoothingFilter
All Implemented Interfaces:
DEMFilter

public class SmoothingFilter
extends Object
implements DEMFilter

Uses a kernel based approach to filter the given dem raster tile. First the standard deviation on the raster tile is calculated, after that a kernel will move over all samples and determines the distance of the local standard deviation to the global std. and determines which samples are used to (re)calculate the height of the sample in dependence to it's neighbours.

Version:
$Revision: 21790 $, $Date: 2010-01-06 17:27:43 +0100 (Wed, 06 Jan 2010) $
Author:
Andreas Poth, Rutger Bezema, last edited by: $Author: rbezema $

Constructor Summary
SmoothingFilter(int size, float stdevCorr, SimpleRaster procesRaster)
           
 
Method Summary
 SimpleRaster applyFilter()
          Starts filtering the raster.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

SmoothingFilter

public SmoothingFilter(int size,
                       float stdevCorr,
                       SimpleRaster procesRaster)
Parameters:
size - kernel size must be an odd value > 1
stdevCorr - correction factor for standard deviation of kernel field. A higher value will cause better saving of edges but also less error correction
procesRaster - needed to create a writeable raster and some no data values.
Throws:
IllegalArgumentException - if the size is even and not odd.
Method Detail

applyFilter

public SimpleRaster applyFilter()
Description copied from interface: DEMFilter
Starts filtering the raster.

Specified by:
applyFilter in interface DEMFilter
Returns:
the filtered raster tile


Copyright © 2011. All Rights Reserved.