3D Modelling of Organs

[VegetableSausage]

Hi all,

I am a new user to Photoscan so still learning. At the moment, I am trying to create 3D models of human organs for a project at uni and can't seem to get a complete 360 model of the organ.

At the moment, I am trying to model a brain that has been cut in half. I have placed it on its cut side and took photos all around it. What I am struggling to get is the side that it is laying on. When I flip it over and take photos and attempt to merge the chunks, it doesn't work. Is there a way to capture or model the brain in such a way that I can get a clean 360 degree model without damaging the organ ? They're very delicate and so far I am just placing them on a white tray.

Any tips or advice would be very appreciated

[James]

I'm not familiar with handling human brains so this is guesswork, but:

Is the half brain rigid or 'squishy'? If squishy then turning it over will distort it so it will no longer be the same shape as when you scanned the other side.

Is it shiny? If so then turning it over may cause it's appearance to change considerably i.e. if it was lit from above previously then turning it over means it is now lit from the opposite direction and the reflections/shadows will have all 'flipped'.

If it's not squishy, and not shiny, then probably you just aren't getting enough overlap between your upside down and upside up photos.

Do you have an image dataset you can share? Really want to see it now!

[Alexey Pasumansky]

Hello VegetableSausage,

You need to mask out the static background to be able to align all images in the single chunk. In principle, you can create the rough mesh models in the separate chunks for "top" and "bottom" subsets, then generate masks from model and then drag and drop all images from both chunks to the single one and run the alignment using "Constrain features by mask" option.
But of course, like James has already mentioned, it's necessary to have good overlap between the images from the different subsets.

[VegetableSausage]

@James

The brain is indeed both squishy and shiny and I can see what you mean in those two factors changing its appearance. I have tried my best to get lighting from all directions to reduce the reflections and shadows.

@Alexey

I think this is what I am looking for... Is there a tutorial on how to perform this exactly? I believe my image dataset should have enough overlap!

Thanks for the help so far !