Difference between revisions of "Hugin hdrmerge"

From PanoTools.org Wiki
Jump to: navigation, search
m (Thomas moved page Hdrmerge to Hugin hdrmerge: Fixes program name)
 
Line 1: Line 1:
hdrmerge combines multiple overlapping images into a single HDRI (high dynamic range image). It is useful for combining multiple LDR images into a single HDR image.
+
hdrmerge combines multiple overlapping images into a single HDRI (high dynamic range image). It is useful for combining multiple LDR images into a single HDR image. The input images should be in linear colorspace.
 +
 
 +
There are 2 merge modes available:
 +
* Average: here the average of each pixel is calculated. Over and under exposure pixel are ignored.
 +
* Khan: Here the Khan algorithm is used to reduce or eliminate in best case ghosts (e.g. moving objects in the different images).
 +
 
 +
== Parameters ==
 +
 
 +
* <tt>-o|--output prefix</tt>: sets the filename for the output
 +
* <tt>-v|--verbose</tt>: activates verbose output (for more details during processing).
 +
* <tt>-h|--help</tt>: Shows the help
 +
* <tt>-m mode</tt>: Sets the merge mode. Can be <tt>avg</tt> (default) or <tt>khan</tt>.
 +
 
 +
If the merge mode is average, the follow parameters can be used
 +
*  <tt>-c</tt>: Only consider pixels that are defined in all images.
 +
 
 +
The Khan algorithm can be fine tuned with
 +
* <tt>-i iter</tt>: Sets the number of iterations to execute (default is 4).
 +
* <tt>-s sigma</tt>: Sets standard deviation of Gaussian weighting function (sigma > 0); default: 30.
 +
* <tt>-a set</tt>: Set advanced settings. Possible options are:
 +
** <tt>f</tt>: use gray images for computation. It's about two times faster but it usually returns worse results.
 +
** <tt>g</tt>: use gamma 2.2 correction instead of logarithm
 +
** <tt>m</tt>: do not scale image, NOTE: slows down process
 +
 
 +
[[Category:Software:Hugin]]

Latest revision as of 17:48, 10 July 2017

hdrmerge combines multiple overlapping images into a single HDRI (high dynamic range image). It is useful for combining multiple LDR images into a single HDR image. The input images should be in linear colorspace.

There are 2 merge modes available:

  • Average: here the average of each pixel is calculated. Over and under exposure pixel are ignored.
  • Khan: Here the Khan algorithm is used to reduce or eliminate in best case ghosts (e.g. moving objects in the different images).

Parameters

  • -o|--output prefix: sets the filename for the output
  • -v|--verbose: activates verbose output (for more details during processing).
  • -h|--help: Shows the help
  • -m mode: Sets the merge mode. Can be avg (default) or khan.

If the merge mode is average, the follow parameters can be used

  • -c: Only consider pixels that are defined in all images.

The Khan algorithm can be fine tuned with

  • -i iter: Sets the number of iterations to execute (default is 4).
  • -s sigma: Sets standard deviation of Gaussian weighting function (sigma > 0); default: 30.
  • -a set: Set advanced settings. Possible options are:
    • f: use gray images for computation. It's about two times faster but it usually returns worse results.
    • g: use gamma 2.2 correction instead of logarithm
    • m: do not scale image, NOTE: slows down process