No announcement yet.

Combine top and bottom?

  • Filter
  • Time
  • Show
Clear All
new posts

  • Combine top and bottom?

    I have a plastic enclosure which I painted with various random colors of paint to create a texture on the plastic surface. I took a bunch of pictures of the enclosure, starting with the "fly around" of the bottom side. I then flipped the enclosure over to do the "fly around" of the top side. I masked the images and ran the workflow. I was able to get a fairly successful mesh of the bottom and the side walls. However, the top (which, due to the order of picture taking, was now on the 'bottom') did not get reconstructed. Attached are picture examples of the top and bottom sides.

    Is this because I am running the Lite version? It was not clear from the tutorial videos whether putting an object in a different orientation would allow the different sides to eventually be pulled together regardless of the version of the 3DF Zephyr.

    Or is it more likely because I did not have enough of a "fly over" along the sides to allow the algorithm to successfully combine the top with the bottom?

    Thank you!
    Last edited by toybuilder; 2019-08-28, 08:39 AM.

  • #2
    Hi toybuilder,

    the technology is the same between Free/Lite/Pro/Aerial - the same input will always lead to the same output across the different versions.

    From what you say, i think you only took a "fly over" of the top and one "fly over" of the bottom. Likely then (if you have masked) you need a few more pictures of at least one side in order to give zephyr enough overlap on the sides.

    I would probably put it vertically and do a capture orbiting around it (end eventually flip it again if you feel the now other bottom side hasn't been captured well enough as well).

    If you want to share the full dataset (photos and zep) I'll be happy to have a look at it and give you more specific insights!


    • #3
      Dear Andrea,
      Thank you very much! I suspected it was more likely that I did not have enough data! Ok, I have a related question -- If I have to "hold" the item such that part of the case is blocked off (imagine maybe like a clamp or a hand holding the case still), so that say only 50% of the case is visible -- as long as I mask away the non-target item, that would contribute enough data to help form that path from back to front?
      I will add more pictures and see how that goes!


      • #4
        happy to help!

        Yes, absolutely simply mask that clamp (and the background) like you normally would!


        • #5
          Ok, after a bunch more pictures and some experimentation, I was able to get all sides of the enclosure to be captured, aligned. and reconstructed. There were some bulges and holes in the end result which I'm sure are the result of my beginner lighting and camera motion techniques. It's definitely not as easy as one initially wishes/imagines it to be, but this tool is clearly quite powerful, and I can already see how with experience, the results would be wonderful.