Unstructured point cloud to pseudo-structured

[FortisNY]

Hello,

I'm using a FJD Trion P1 Slam scanner, and the output data is one big pointcloud.

I have in mind, since there is possibility from nearly all slam devices to export Pano poses, and pano pictures to make a plugin which could do:

- Take a e57 pointcloud
- Take a csv/txt file with pano poses and locations
- Take a input of sphere radius from pano pose and set max. distance based on this value (cut everythiing behind this radius)
- Export a single e57 files with embeded pano picture, segmented pointcloud (with defined radius)

I've tried put something with ChatGPT, but so far, not working :) What do you think?

[daniel]

With ChatGPT ?!

Well, you'll probably need stronger skills in C++ and Qt. And do it without much AI support considering it's quite 'niche'.

[FortisNY]

Well, yes, my programming skills are not sufficient. I can read and understand code and what's going on in it, but can't code it from ground :) so with the help of AI I did the python script that read the point cloud and export the smaller point clouds based on pano locations and radius. Unfortunately there is no python library that can write e57 with embedded panos and virtual sensor positions. That needs to be done in C++ with origin LibE57 library, but I'm afraid that's behind my skills for now...