Multiple components from large data set of full body

Development Asset Creation
,
1

Hi RC team,

I have a turntable data set of 333 x Canon 5DS images of a full body. Unfortunately even though I have good overlap, somehow I have only managed to get a maximum of 267 images to align so far. - Sadly i end up with around 15 - 20 components.

I can’t post the data set as it is under NDA, but I have tried everything from changing alignment settings as per Milos instructions. 10k/40k, 20k/80k features etc etc, high medium and low image overlap, preselector feature, detector sensitivity etc… Even force component rematch.

Surely with 333 sharp cross polarised canon 5DS images you would have thought RC would manage to compute 1 component?

Background is 18% grey and out of focus so should not be interfering with images. Images shot under flash photography, ISO 320, F10, 100mm macro f2.8. Plenty, plenty of overlap. Any suggestions?

Apologies if I am asking a question that others have asked and had answered already. I am desperate.

Any help would be greatly appreciated.

2

Hi bananajunky

Can you contact me at milos.lukac@capturingreality.com?

3

I’m Guessing that is one camera many shots on a turntable, My guess is that the person is moving.

4

Hi Milos. I will get in touch via mail thank you…

Full body is not human Steven so your guess is incorrect. It is a static object. Do you have any suggestions you can offer?

5

My only guess is that RC is picking up info in the background and confusing the alignment. is it an absolutly plain background, you could try masking the offending images