Aligning components breaking up

Answered

Comments

4 comments

  • Avatar
    ssh4
    Hi Ben.

    You can try to lock cameras in both components and align again.

    But better if next time you will have proper overlap between top and bottom part and will use masks (as alpha chennel for this moment).
    0
    Comment actions Permalink
  • Avatar
    Wishgranter
    Hi Ben
    please make a screenshot so that we can see the CP details ( in how many images they are placed ) and the point placement strategy..
    0
    Comment actions Permalink
  • Avatar
    OzBigBen
    Possibly a bit short on CPs ;) 3 images per CP per component although some only have 2.
    Image
    There are also no more than 2 CPs in any image which might not be the best.

    With the constraints of the lighting setup I could only get the top and sides before turning the book over. Normally I'd photograph underneath as well and mask out the support. All images are masked to just the object. The patterns on the front and back are very similar, with the small picture in the middle being the main difference. I'm not expecting to get a perfect object without doing some mesh editing if I can get the two halves aligned.

    Should I be doing anything with Locked pose group, Lock for Continue?
    0
    Comment actions Permalink
  • Avatar
    Wishgranter
    Hi Ben
    Yup, you need at least 3 images per CP/COMPONENT, if you get say 5-7, it’s OK in most cases. Try to get 3-4 CP per image as this helps ALIGNMENT to estimate data properly.
    A second option would be to increase CP weight from the default 10 to say 20-50. But just after you’ve added CPs and tested the alignment with them, just then change the weight..
    0
    Comment actions Permalink

Please sign in to leave a comment.