Hello,
I'm using photoscan to make dense 3D model of a complex environment (an underwater rock with cavities) and i'm facing huge time and CPU consumming process.
My original photoscan project is made of a chunk of 5100 aligned photographs (taken from an AVT Prosilica GT660 6576 × 4384 camera:
https://www.alliedvision.com/en/products/cameras/detail/6600.html)
When i'm trying to make dense point cloud of my chunk, the processing is quite long (90 hours already passed, 19% done and still 359 hours to go).
The set-up of the dense point cloud computatio is:
Quality: Medium
Depth filtering: agressive
* The computer that run photoscan is a super-micro type with 2 xeon processors, 512 Gb of RAM and a Geforce GTX 980 graphic card.
The depth map creation step was long bu reasonable (40 hours) but now, the filtering of depth maps is started from 90 hours and is achieved at 19% with a remaining time of 350 hours.
Is anyone can explain why the densification time is so long ?
Do i set a bad parameter ?
Thanks a lot,
Julien