Capturing a Coin and Merging Chunks

[chadfx]

Hello again,

I'm trying to find an elegant way to capture a small coin, and have yet to find a reliable method that allows me to get the entire piece as a complete clean object.

If I shoot around it placed flat on each side and create two chunks in PS, I cannot get enough overlapping points to get the two chunks to merge reliably. This makes sense, as there is very little detail along the coin edges, and it's usually very uniform in appearance.

If I shoot it mounted vertically, I can usually get a decent build of the whole coin, except that I have to secure it somehow to keep it in that vertical position while I shoot it, and I would also like to see that part of the coin that is bearing the support. I have tried to shoot the coin vertically in two passes with the mounting point moved to a different location on the coin for the second pass; and I can get each chunk to build correctly on their own. But I cannot find a way to get PS to align the chunks correctly so that the second chunk with 'fill in the gap' created by the mounting point from the first chunk. Usually it still aligns the two chunks in the exact same shooting positions. I have tried masking out the mounting point in both chunks, but that doesn't seem to align at all.

I am only using the standard edition, so I do not have the option of defining correspondence points between the two chunks. But I would have thought that PS could still recognize enough overlapping information between the two chunks to figure out the alignment.

Are there any ideas out there on how to get this to work? I am stumped at this point.

Thanks! -Chad

[Wishgranter]

hmmmmm, try use newspaper and glue the "10 halier" on it, then just shoot around it. So you have ONE chunk..... or some "milimeterpapier" what can get decent and precise align, then just generate mesh...

[chadfx]

Thanks for the suggestion, Wishgranter. I have been able to do the shoot around technique if I lay the coin flat on a surface. I'm not sure if I am understanding how your newspaper trick would help to capture both sides of the coin to become one final object.

Here's a rough example of a coin captured vertically by using some small holding clip at the bottom. But I want to somehow get that bottom part as well, even if I have to merge it in as a separate chunk...I just can't seem to be able to get it to work.
 
http://sketchfab.com/show/fx7oUjg223LW6SQ3ixY9T6Wb34X

If I shoot both sides 'flat' against a surface, I can get each side to build correctly. But PS cannot find enough common points to align/merge the chunks, and the cameras from both chunks are in the same orientation after the align...so I don't think they can help with the aligning of the two chunks.

[meshmaster]

Thanks for the suggestion, Wishgranter. I have been able to do the shoot around technique if I lay the coin flat on a surface. I'm not sure if I am understanding how your newspaper trick would help to capture both sides of the coin to become one final object.

Here's a rough example of a coin captured vertically by using some small holding clip at the bottom. But I want to somehow get that bottom part as well, even if I have to merge it in as a separate chunk...I just can't seem to be able to get it to work.
 
http://sketchfab.com/show/fx7oUjg223LW6SQ3ixY9T6Wb34X

If I shoot both sides 'flat' against a surface, I can get each side to build correctly. But PS cannot find enough common points to align/merge the chunks, and the cameras from both chunks are in the same orientation after the align...so I don't think they can help with the aligning of the two chunks.

Here is how I do it...

Put a few, very small bits of clay... little balls along the edge of the coin.  You can then use them as alignment features as you should be able to see them from both sides of the object.

Of course you will need to edit the geometry and remove them afterward and fill the holes, but it works a treat... I've done this myself many times.

Cheers,

Joe

[chadfx]

Thanks Joe, that's a pretty clever trick...sort of like adding 'training wheels' to it. I'm hoping to avoid the cleanup stage, but maybe that's just something I'll have to resign myself to.

Are you then capturing each side laying flat down and merging the two sides? Or are you shooting with it propped up and just using the extensions to aid in the alignment/solve?

I was able to merge two flat chunks in Meshlab, as they have an alignment feature between to objects loaded as layers...but when I tried to merge the two aligned layers, the textures broke.

Cheers, -Chad

[meshmaster]

Thanks Joe, that's a pretty clever trick...sort of like adding 'training wheels' to it. I'm hoping to avoid the cleanup stage, but maybe that's just something I'll have to resign myself to.

Are you then capturing each side laying flat down and merging the two sides? Or are you shooting with it propped up and just using the extensions to aid in the alignment/solve?

I was able to merge two flat chunks in Meshlab, as they have an alignment feature between to objects loaded as layers...but when I tried to merge the two aligned layers, the textures broke.

Cheers, -Chad

I would shoot it standing up, rather than laying flat then using the extensions to aid in the alignment/solve.

Yes, to be honest I'd most likely do the alignment in a 3rd party application myself but as you point out, the textures could be problematic depending upon what you are using to do the job.

Objects with thin edges and low features are always a challenge -- always!

Keep at it.

Cheers,

Joe

[meshmaster]

A few more tips...

I would not put the clay balls in a "uniform" distribution around the edge of the coin, rather I would stagger them.  Put a couple close to each other, another further away and so on... make them different sizes too, this will make them easier to identify.

I would likely do 2 scans using this method, however, on the second scan I would likely rotate the coin a bit from it's original position (this is so that the area currently occluded by the alligator clip would be revealed). I would also move the balls around the edge to a new location, making sure that they are NOT in the same spots as during the first scan.

Then once you have 2 scans, you can export, and align when you remove the balls the hole filling will be easier because you should have all the areas of the coin that you need.   It's really just a matter of aligning coin scan 1 to coin scan 2 and your done.

This is the method I always use for for such objects and it works great.

Let me know how you get on.

Cheers,

Joe

[chadfx]

Yeah, uniformity and PS do not mix well. ;-)

Thanks Joe, good tips all around. I was essentially attempting this idea (rotate coin in clip for 2nd chunk), but couldn't get the align and merge to work. I was hoping to delete the pieces where the clips are and just merge the 'clean' sections. Wouldn't that be handy? No additional geom or texture cleanup needed! Ah well, maybe it just isn't possible.

Actually, I wonder if keeping the clay helpers in the same positions for both chunks would help with the auto align? But I'd still be left with reconstructing the clay covered bits.

Have you tried using colored clay instead of varying the pattern/sizes? Maybe it's time to hit the art supply store and get some Sculpey.

Cheers, -Chad

[airmap3d]

I really have no idea if this would work or not but...

Could you try having it suspended by something like a human hair held on with a tiny dot of super-glue?  That way you could get all around it in one go with only a very small blemish to fix up after processing?

Just a thought!

[chadfx]

Just to say that it was "captured by the width of a human hair" seems to beg for it to be attempted, no? ;-)
I applaud your creativity...although for a lot of objects, the notion of super glue would not bode well

My wife thought about trying to suspend it in mid air with magnets in a repel orientation, but I think that would depend on what material the coin was composed of and difficult to keep stable.

Definitely some sort of 'minimal contact' setup would be great...it just needs to be relatively stable as well and hopefully not interfere with any lighting setup.

[James]

Another creative idea, shoot with both the camera and coin in freefall. I'm not sure about the specifics of the release mechanism, which would require some level of sophistication, but if you can allow a 4 second drop (about 80 meters) then you could get several shots of the coin rotating in free space.

Repeat as necessary

[tincansassoc]

Your clamping setup is sufficient. I have done similar objects like this all in one chunk. Zoom-in more and take a lot more photos. If you do it right there will be no cleanup work required. No merging chunks required.

[chadfx]

Thanks TinCan, so just to be clear...

If I capture sufficiently high detail (more photos, focus stacking, etc) and use a clamp to hold the object in two different positions for two series of photos of the object; then PS should be able to align everything in one chunk?

How do I get rid of the two sections of the clamp to reveal the clean object underneath?

[tincansassoc]

Just move the clamp and take more photos where the clamp was. Photoscan will fill in the missing info as long as the clamp is masked out.

[chadfx]

Sounds great, that's exactly what I was hoping it would do. Maybe I just don't have enough coverage and detail to cover it, then.

I'll try a more detailed re-shoot and apply focus stacking to see if I can get this to work.

Thanks again - Chad