Forum

Author Topic: Hello, I am new to photogrammetry and I have lots of very specific questions  (Read 2249 times)

tymal77

  • Newbie
  • *
  • Posts: 4
    • View Profile
I have access to metashape pro and was hoping if I could get some help with a few issues I've had since I've started. I'll try and upload photos to go along with this. I'm using a Google pixel 2 with focus and contrast lock to take my photos, and a custom built pc with an i5 and an AMD Radeon R9 390 series as my PC.

Firstly, I've watched many videos on using this software and have found that the meshes and dense clouds people are producing are much more complete while mine have been very broken up and contain lots of holes. I'm on my fourth try now and my results have improved each time but I still feel like I'm doing something fundamentally wrong.

Secondly, my GPU does not appear in my GPU selection preferences, I'll post a photo of what it looks like along with this. I'm assuming this has affected my end result in some way.



Thank you for your time to anyone that can help. I know I probably need better photography equipment but I'm trying to get the best results with what I have right now. I also want to note that I have experience in zbrush and blender, and while I'm just focusing on objects right now I'd like to move on to larger landscapes about an acre to 2 acres in size.

If you'd like access to any of my files please let me know and I'll see what I can do.
« Last Edit: May 28, 2019, 06:48:45 PM by Alexey Pasumansky »

Dave Martin

  • Full Member
  • ***
  • Posts: 170
    • View Profile
Tymal77,
As a first step, I would suggest you share one of your source images and also a screenshot of your resultant model.
Dave

pbourke

  • Jr. Member
  • **
  • Posts: 58
    • View Profile
    • Personal/professional web site
A couple of comments

- I wouldn't worry about the GPU, my understanding is it won't make any difference from a model quality perspective, only the time taken.

- With regard to the quality of the results, what are you trying to reconstruct? There are well known/understood problem cases. For example smooth featureless surfaces, highly specular or reflective surfaces.

- And lastly and more speculative .... I've had mixed success with mobile phones. One of the issues I imagine is that phones these days are doing all sorts of processing on the images, often without you knowing or being able to do anything about it. I wonder if this processing which can include global changes can degrade coherence of what would otherwise be good across image feature points. Even with my DLSR capture I turn image processing/enhancements (or whatever the particular manufacturer calls it) off.
Photographic reconstruction portfolio
http://paulbourke.net/reconstruction/portfolio/