Author Topic: Processing large numbers of images - advice needed!  (Read 647 times)


  • Newbie
  • *
  • Posts: 13
    • View Profile
Processing large numbers of images - advice needed!
« on: February 14, 2022, 02:21:40 AM »

We are using a 6 camera rig to survey reefs over large areas. Each survey is around 0.5 Ha and contains 60-90K images.

We are struggling to find a computing platform (in-house or cloud based) to process these and I was wondering if anyone is routinely processing projects of this size and if so:

  • What is your workflow - that is what Metashape settings do you use for the basic processing through to an Ortho-mosaic?
  • What hardware do you have or use - i.e. processors / graphics cards, etc
  • Do you do this in house with your own resources or via a cloud provider such as AWS?

At the moment we align all of the images in one go and then chunk them for the depth maps and mesh generation but this takes a long time and often slows down due to memory and so on. So we are looking for some ideas about how this is being done by others.

Happy to provide exactly what we are doing now if that helps but thought I would ask more general questions to see what is working for others.




  • Jr. Member
  • **
  • Posts: 59
    • View Profile
    • AccuPixel Ltd - Dealer and Training Centre
Re: Processing large numbers of images - advice needed!
« Reply #1 on: February 14, 2022, 02:15:16 PM »
We have done some consulting work with the US National Parks Service such as this:

Happy to discuss further, if interested? Please use the Contact Us page:
Agisoft endorsed online Metashape training - see: