Forum

Author Topic: automated focus stacking?  (Read 3978 times)

driftertravel

  • Newbie
  • *
  • Posts: 46
    • View Profile
automated focus stacking?
« on: December 24, 2014, 10:46:23 PM »
Just wondering, if it were possible to set up an automated focus stacking system for a photogrammetry rig would this massively increase the usable data present in an image, or would it be just as easy to run the separately focused images through photoscan, or would it otherwise mess up the depth calculations?

Alexey Pasumansky

  • Agisoft Technical Support
  • Hero Member
  • *****
  • Posts: 14813
    • View Profile
Re: automated focus stacking?
« Reply #1 on: December 25, 2014, 11:28:43 AM »
Hello driftertravel,

On this forum you can find some topics related to the focus stacking.

From our side we suggest to use external application to perform this operation and load already stacked images to PhotoScan. However, in the version 1.1 you can group all the images taken from the same camera position to the Camera Station instance in the Workspace pane and proceed to the image alignment operation. But due to blurred parts on the images this way in not recommended as it will affect every reconstruction step.
Best regards,
Alexey Pasumansky,
Agisoft LLC

JohnyJoe

  • Jr. Member
  • **
  • Posts: 62
    • View Profile
Can photoscan autofocus-stack images withinh itself?
« Reply #2 on: January 03, 2015, 08:47:03 PM »
Hello
Can photoscan automatically "focus stack images" inside the application itself during the processing, what i mean mainly:

1) If i take 5 exactly the same shots, just with the focus varying, i should use special app like helicon focus for stackign, but what if i put these images directly into photoscan, will it be able to take only the sharp section of each of the images during processing into account? For example PS recognizes that only the lower third of the screen on image 1 is shaprt so it will "work" with it (for sparse, dense, mesh and texture generation), for the 2nd half of the image, PS recognizes that it should work with the 2nd image etc. So basically focus stacking (automatic) inside photoscan itself, is it possible?

2) And second question, if i take for example 24 shots of an object, and some photos some area of the object are out of focus, but on the other ones they are not, can photoscan recognize this and use it? For example on 1st photo only a head of dog is in focus so it takes this information, the tail of dog on this photo is out of focus, but the tail is shaprt and INfocus on the 8th image for example, where vice-versa the head is blurred (out of focus), photoscan can recognize this and take the "head information" (for clouds, meshes, textures etc.) mainly from the 1st photo and the information about the tail mainly from the 8 photo, am i right?

Thank you