Author Topic: Auto select photograph of greatest image quality from a set  (Read 4580 times)

Offline pshrutpark

  • Newcomer
  • *
  • Posts: 48
    • View Profile
    • Harris Photography
I often take three photographs in quick succession if I'm shooting hand held and the image quality is especially important. I then compare the three images and delete the two images that are least clear.

With many sets of three images this is quite time consuming.

It has occurred to me that this might be done semi-automatically as a function in PM, probably based on image contrast analysis.

One could select photographs of the set and choose "Select Best" from the menu. The best image is retained and the others deleted.

Does this function appeal to others?

Peter

 

Offline vAfotoriporter

  • Uber Member
  • ******
  • Posts: 1046
    • View Profile
    • Attila Volgyi photojournalist
Re: Auto select photograph of greatest image quality from a set
« Reply #1 on: July 13, 2012, 12:58:55 AM »
I often take three photographs in quick succession if I'm shooting hand held and the image quality is especially important. I then compare the three images and delete the two images that are least clear.

With many sets of three images this is quite time consuming.

It has occurred to me that this might be done semi-automatically as a function in PM, probably based on image contrast analysis.

One could select photographs of the set and choose "Select Best" from the menu. The best image is retained and the others deleted.

Does this function appeal to others?

Peter

I would rather not let the choosing to any automated feature however good an algorythm can be. Choosing manually between images is the key of PM and I think developing the Artificial Intelligence that would be capable of doing what you request would take fare more energy away from developing other more suited and needed features.
Working on Mac, OSX, iOS and with some Canons.
Allways shooting RAW.

http://www.volgyiattila.hu

Offline Sven

  • Uber Member
  • ******
  • Posts: 1053
    • View Profile
Re: Auto select photograph of greatest image quality from a set
« Reply #2 on: July 13, 2012, 01:14:02 AM »
Hi!

If you want to check the "sharp" areas of an image (RAW only) give Capture One from Phase One a try.
Areas shown in the light green are the sharp ones.

Sven

[attachment deleted by admin]
After 5 years of absence I restarted the photography.

Offline pshrutpark

  • Newcomer
  • *
  • Posts: 48
    • View Profile
    • Harris Photography
Re: Auto select photograph of greatest image quality from a set
« Reply #3 on: July 13, 2012, 02:19:31 AM »
Thanks for your comments.

The idea is to eliminate blurred images from a set of otherwise identical images, intentionally taken in the hope that one would be sharper. Not to generally select the best images from a photoshoot.

Probably AI would be more than required to do that. I imagine contrast analysis, maybe similar to routines behind contrast based auto-focus, would work.

Definitely don't want to detract from the development of V5 and Catalog!! 

In Capture One does the green indicate sharp  versus less sharp or are there shades of green? In the scenario I describe what would I expect to see, in Capture One, if one image is sharper than the others?

Again thanks.

Peter
« Last Edit: July 13, 2012, 02:26:01 AM by pshrutpark »

Offline Sven

  • Uber Member
  • ******
  • Posts: 1053
    • View Profile
Re: Auto select photograph of greatest image quality from a set
« Reply #4 on: July 13, 2012, 02:24:09 AM »
Peter,

you can adjust the threshold for the sharpness.
C1 does only display the sharp areas that are (more or less) in focus.

Sven
After 5 years of absence I restarted the photography.

Offline pshrutpark

  • Newcomer
  • *
  • Posts: 48
    • View Profile
    • Harris Photography
Re: Auto select photograph of greatest image quality from a set
« Reply #5 on: July 13, 2012, 02:28:17 AM »
Peter,

you can adjust the threshold for the sharpness.
C1 does only display the sharp areas that are (more or less) in focus.

Sven

OK. Thanks.