EDGEBLUR

Blur (smooth) edges along an elevation jump


EnvironmentsPYTHON :: EASI
Quick linksDescription :: Parameters :: Parameter descriptions :: Details :: Example :: Related

Back to top

Description


EDGEBLUR is used typically to blur harsh edges in a digital surface model (DSM) (or true) orthorectified-image mosaic along elevation jumps, such as those along building or tree edges.
Back to top

Parameters


edgeblur(filedem, dbec, filo, dbimage, demfilt)

Name Type Caption Length Value range
FILEDEM * str Geocoded DSM to use to find elevation jumps 1 -    
DBEC * List[int] Input DSM channel 1 - 1  
FILO * str File name of image (mosaic) to process 1 -    
DBIMAGE List[int] Image channels to process 0 - 8  
DEMFILT * str Filter level 1 -    

* Required parameter
Back to top

Parameter descriptions

FILEDEM

The path and file name of the geocoded DSM file to use to produce the true-orthorectified or DSM-orthorectified image mosaic.

The file must be of the same resolution and projection as the image mosaic specified for FILO, though the extents can differ.

DBEC

The input channel that contains the DSM or digital elevation model (DEM) to examine for large elevation changes.

FILO

The path and file name of an existing output imagery file.

The files specified for FILEDEM and FILO be of the same resolution and projection, though the extents can differ.

DBIMAGE

The imagery channels in which to blur edges along an elevation jump.

If no value is specified for this parameter, all bands in the file are processed. The imagery channels will be overwritten.

DEMFILT

The level of filtering to apply in edge blurring.

The available values are as follows:
Back to top

Details

True-orthorectified (or DSM-orthorectified) airphoto (or UAV) mosaics often have harsh, unnatural-looking, edges along the edges of buildings and and trees where there are large, sharp, elevation changes. By applying a blurring filter only along these edges a more pleasing natural appearance is achieved and small errors become less noticeable. The vast majority of the image pixels remains unchanged.

EDGEBLUR applies a small (3 x 3) Gaussian smoothing filter on the imagery data (specified in FILO and DBIMAGE), as guided by large elevation jumps in the DSM file (FILEDEM and DBEC). A large elevation jump is assumed to be four times the imagery resolution or 1.5 meters, whichever is greater.

Typically, the DEM file specified in FILEDEM is a DSM generated from a stereo-airphoto project and is the same one used in generating the true-orthorectified imagery. The file specified for FILEDEM and FILO must be of the same resolution and projection, though the extents can differ.

The strength of the filtering (blurring) applied along the edges can be specified in DEMFILT. LOW applies a very light, 3 x 3 Gaussian blur, MEDIUM applies a heavier, 3 x 3 Gaussian blur, and HIGH applies the medium filter twice.

The typical workflow is as follows:
  1. Create a DSM mosaic by running the EPIPOLAR, EPIPOLARDSM, GEOCODEDDEM, and DSMMERGE algorithms
  2. With the DSM mosaic you created, run TORTHO to create true-orthorectified imagery
  3. Mosaic the true-orthorectified images by running MOSPREP and MOSRUN
  4. Blur the edges of the true-orthorectified mosaic by running EDGEBLUR
Back to top

Example

You have a RGB true-orthorectified image mosaic (image_mosaic.pix) and the DSM-elevation file (dsm.pix) used to produce the true-orthorectified images. Both files must be of the same projection and resolution.

from pci.edgeblur import edgeblur

filedem  = r"dsm.pix"  
dbec     = [1]
filo     = r"image_mosaic.pix"
dbimage  = [1,2,3]
demfilt  = r"medium"

edgeblur(filedem, dbec, filo, dbimage, demfilt)

© PCI Geomatics Enterprises, Inc.®, 2026. All rights reserved.