Point Cloud Vertical Details

Comments

8 comments

  • Avatar
    Wallace Wainhouse CR

    Hi, 

    This looks like a very unusual result, there also doesn't seem to be much geometry on the trees. It would be interesting to see the source images. Did you use the GCP in the image, if so what kind of errors did you get on it? You might be able to get the post back by adding some control points to it.

    0
    Comment actions Permalink
  • Avatar
    Frank Zwick 1

    Hi,

    maybe the height of the ReconstrRegion is too low?!

    1
    Comment actions Permalink
  • Avatar
    Wallace Wainhouse CR

    Hi,

    I think Frank Zwick might be right, it does look like everything has been sliced off at the same height. Thank you I didn't think of that.

    0
    Comment actions Permalink
  • Avatar
    ASRich327

    Thank you for your replies. I do not see how I can reply to each one individually. Wallace, I did not use GCPs. I wanted to compare apples to apples. iWitness Pro V4 and Reality did not reconstruct the pole. 3DF Zephyr and Pix4D DID reconstruct the pole. In all four cases, no GCPs were used. I am not a photogrammetry expert, but I have been using photogrammetry for 20 years. I do not see how GCPs can affect the reconstruction region. I use GCPs for accuracy. I can be wrong. As I said--I am not a photogrammetry expert. If you know how GCPs can affect my pole getting reconstructed I would love to learn from you! Also, how did you tag Frank in your reply?

    Frank: Nice call on the reconstruction region, but it was set just above the height of the house, so I would expect that at least some of the pole would be reconstructed? I raised the region to well above the height of the pole--same results.

    I will now add the GCPs and see if that works. I will let you know. Again, THANK YOU both for your suggestions!

     

    Andy

    0
    Comment actions Permalink
  • Avatar
    ASRich327

    Okay. So I spent 2 hours adding in GCPs (It actually only took an hour, but I forgot to save and Reality Capture crashed when I aligned--so I had to redo all 6 GCPs).

    I had 6 GCPs in the scene. They were measured with a total station. It did not help at all. Still no pole. I could not add control points to the top of the pole. There was over 2 PX error. My guess is that my GCPs only had an elevation difference of 2 feet. Not enough to lock in the z-axis.

    Thanks for the great suggestions, though!

    Andy

    0
    Comment actions Permalink
  • Avatar
    ASRich327

    Oh, and the GCPs were placed in at least 5 images, and some up to 12 images.

    0
    Comment actions Permalink
  • Avatar
    Wallace Wainhouse CR

    Hi,

    Sorry I should not have mentioned the GCPs. I was only curious about the error range on them if you had already added them. What I was actually suggesting was to try adding some control points on the post itself. If you like I can take a look at the project, or perhaps just send me the images for this specific area, and I can see if I get a similar result. support@capturingreality.com

    Also you can tag, just by copying and pasting someone's name

    0
    Comment actions Permalink
  • Avatar
    Wallace Wainhouse CR

    Hi,

    also please could you check if you have a clipping box active?, and if so clear it.

    0
    Comment actions Permalink

Please sign in to leave a comment.