- This topic has 0 replies, 1 voice, and was last updated 1 day, 19 hours ago by Macroscopic Solutions.
-
AuthorPosts
-
December 20, 2024 at 12:21 pm #21662
Dear Agistoft.
I have a very ambitious project where we sent plant roots to space aboard Polaris Dawn and are 3D modeling the root structures. We were very limited by the containers we used. They had to be plastic, square and frozen upon return prior to gene sequencing. I have over 118 models I need to generate.
I’m getting good alignments of the cuvette containers (Image 1). However, the root structures are being misaligned because of optical refraction when photographed through a planar surface from an oblique angle. I have to proceed with a dense point cloud method because the stereo method will not work because of this refraction problem (Image 2).
Image 1.
Image 2.
After a quick crop to isolate the root structure. I now need to proceed with a method that allows me to combine the four prominent structures into one accurate and unifying structure (Image 3). I have tried markers, but because each structure to twist. Im also batch processing over a dozen samples for my initial alignment, so I want to keep my workspace (Image 4) fairly clean without having too many duplicated chunks. How do you advise I proceed. What steps should I take to achieve one unified root structure considering the challenges I am confronted with?
Image 3.
I did a practice run over a year ago and published this here: https://www.youtube.com/watch?v=1zHiX2zy6a4
Some models to come together very nicely on their own, the majority of them do not.
You can see an example of a dense point cloud that merged exceptionally well below at Image 5.
Image 5.
- This topic was modified 1 day, 19 hours ago by Macroscopic Solutions.
- This topic was modified 1 day, 16 hours ago by Macroscopic Solutions.
-
AuthorPosts
- You must be logged in to reply to this topic.