Forum

Author Topic: Camera-based chunk alignment/ merging chunks  (Read 4087 times)

bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Camera-based chunk alignment/ merging chunks
« on: November 27, 2015, 01:58:26 PM »
Hi All

I'm working on a larger project at the moment and am trying to get a better practical understanding of aligning and merging chunks.

Original project (well the first 1/3 of it) is ~1200 cameras which I've split into 8 chunks of 200.  The model essentially follows a winding path, so each chunk has an overlap of about 50 cameras with the previous chunk for alignment.  I've done a couple of quick tests and the alignment works pretty well but I'm wondering what the best process is for generating the dense cloud.

Is it better to create a dense cloud of each chunk and merge them, or merge the aligned chunks and then create the dense cloud? Is either quicker?

If I create dense clouds of each chunk, there will be sections at the end of each chunk of increasing noise.  Each join between chunks would then have two sections of increased noise either side of the join which wouldn't be there if a dense cloud were created from all of the images at once.... but is doing the dense cloud with all images practical (time vs. quality)

Bits of the project are here: https://sketchfab.com/uomdigitisation/folders/william-ricketts-sanctuary  I'm attempting to make a point cloud of the entire reserve including detailed sculptures.  Early tests are looking very promising but it's my first attempt at a significantly larger point cloud.  Any suggestions/advice gratefully accepted.

PS. Loving the 1.2 update.
« Last Edit: November 29, 2015, 03:00:14 AM by bigben »

bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Re: Camera-based chunk alignment/ merging chunks
« Reply #1 on: November 30, 2015, 12:20:18 AM »
Dense point clouds finally finished and merged. There are a few areas where the alignment wasn't quite right, screengrab being the most noticeable area.  [edit]Point cloud taken down as part of a clean up[/edit]

20mp images, full frame fisheye
250 cameras per chunk, 50 cameras overlap between chunks
Medium alignment (fixed lens parameters), medium dense cloud

The amount of overlap would be roughly 20m on a straight section of path.

Any suggestions?

Would merging chunks and then optimising cameras be better? I'm guessing not, as the overlap cameras are duplicated when merged.
« Last Edit: February 13, 2016, 06:39:04 AM by bigben »

temporalista

  • Newbie
  • *
  • Posts: 1
    • View Profile
Re: Camera-based chunk alignment/ merging chunks
« Reply #2 on: February 12, 2016, 07:09:21 PM »
Hi bigben.

I have the same question, i.e. if it is better to create separate dense clouds and align them at the end or otherwise.

Have you new results?


bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Re: Camera-based chunk alignment/ merging chunks
« Reply #3 on: February 13, 2016, 06:51:37 AM »
Well it took a while, but I created a point cloud from all of the images in this set which took a week to align and a few days to produce a dense cloud.  The result was, as expected, pretty good. http://files.digitisation.unimelb.edu.au/potree/pointclouds/ricketts-test2.html  but this is about as long as I'm prepared to have a machine processing a single chunk.  I had thought of aligning one chunk, exporting the last few cameras and importing them into a new chunk with fixed positions before adding the extra images.  Given that I'm using fisheye lenses and the subject is relatively complex, a small shift in camera position can create very noticeable differences. 

Still interested to hear any suggestions.