I’ve got a Component with some 164 cameras aligned, had created various tie-ins to build up this bigger component from about six smaller ones. I have just a couple of images in the sequence that didn’t align, and when I add tie-ins to these and overlapping images in the aligned set, then running alignment, I’m surprised to get 4-5 Components with a combined number of maybe 16 cameras, the vast majority of the previously aligned images not showing up in these Components. The Control Points have reasonably low residuals, no yellow warning signs. How can this be and what can be done?
Here’s the scene with all the cameras but this small handful:
Please Shift Run RealityCapture for reset settings to default. And later don’t change settings without strong understanding.
As i can see you set extremely low Max Features per image. That can be main reason of failed triangulation on Low overlapped images.
If you ave good sharp images even with low overlap RC can align all camera as one component from first attempts.
But as usual more images (better overlap) can help more.
Thanks for lifting a thorn! I couldn’t get squat to work, even on smaller batches that worked before, how obvious now. I had actually upped the max features to 80,000 to overcome poor alignment in an area where the cave wouldn’t grant me proper access to the angles I needed, and in resetting I made the typo, leaving out a zero. Very well, many thanks.