Forum

Author Topic: Laser scan preserving position  (Read 4363 times)

4dtehnik

  • Newbie
  • *
  • Posts: 1
    • View Profile
Laser scan preserving position
« on: October 23, 2023, 11:08:13 PM »
Hi,
i done some laser scan and photo aligment and results are great. I use this topic as guide:
https://agisoft.freshdesk.com/support/solutions/articles/31000168474-terrestrial-laser-scans-processing-in-metashape-2-0-0
But,is there some kind way to preserve laser scan position, because when aligment is over, position of laser scan isnt on origin place?
greetings

Tas

  • Newbie
  • *
  • Posts: 49
    • View Profile
Re: Laser scan preserving position
« Reply #1 on: November 02, 2023, 05:24:47 PM »
+1  --  I'm also looking for a solution to this question, as I'm trying to pitch Agisoft Metashape as a solution for combining laser scan and image data. I've personally found the Metashape's alignments to be superior to alternatives, but I've yet to solve the issue of maintaining laser scan coordinates.

My current understanding is as follows (based on Resources/Research list below and dozens of other forum threads)
  • The relative positions of grouped laser scans can be locked.
  • To align images and laser scans, the absolute position of a laser scan (or laser scan group) must be released.
  • By default, precedent appears to be given the photos for absolute positioning - i.e., the laser scan group will be shifted to the camera positions during alignment. I've run a number of trials by altering the accuracy settings of laser scan positions through the reference pane interface, but I haven't had any luck in forcing the absolute position of the laser scan to be fixed. I'll likely start another thread to get into greater detail.
  • After aligning the images and laser scans, there's the question of whether there is a streamlined way to translate & rotate the chunk position so that the laser scan group aligns with its original position. Again, I've run a number of trials without success, and it might be a topic for another thread.
  • In short, I started by trying to use the Python script in the second link listed below, continued with an embarrassing amount of troubleshooting, and ended by trying to write my own script to perform what's essentially an orthogonal Procrustes transformation. The intent was to define the 4x4 transformation matrix that would best align the transformation matrix of one shifted (aligned) laser scan and the same laser scan re-imported in the correct absolute location. Then the transformation matrix would be applied to shift/rotate the position of the entire chunk. I'm working entirely in local coordinates, so I expect that there isn't an issue with coordinate system transformations. I'm not proficient with Python, however, and I gave up when I couldn't successfully load an external Python module into Python (NumPy).

Resources/Research

Any response, input, and/or guidance is greatly appreciated! I'll try to make time to document my trials in a se
« Last Edit: November 02, 2023, 05:26:46 PM by Tas »

Tas

  • Newbie
  • *
  • Posts: 49
    • View Profile
Re: Laser scan preserving position
« Reply #2 on: November 04, 2023, 01:03:46 AM »
Follow-up --  I documented a workaround here: https://www.agisoft.com/forum/index.php?topic=15274.msg68811#msg68811