Hi all!
I require your help regarding a project of mine. I don’t expect to have it done for me; While I’d entertain the idea, I’d also like to do this project “side by side” or with a mentor. I’ve been trying to get this solved for 3 or 4 years now, and it’s killing me. I DREAM about it.
Description:
I’m trying to figure out the best setup for building a robot to help me “scan” artwork using a DSLR. I’ve looked over tons of things, but for one reason or another, I just can’t find a solution. And no, a literal scanner won’t work, becasue the art can’t touch the glass. 
I need something that will allow me to adjust the overlap of images the camera takes, which would allow me to adjust how high a resolution the final stitched image would be. I’m pretty well at a loss. I need to check off:
Hardware
… I’m not sure, but I need it to be able to:
- have a 3-foot radius resulting in a 6 foot up/down/left/fright field
- mount the camera & plug into it for controlled operations
- maintain “Z” distance from the artwork being digitized
- be stable: in some images, I may be required to use a macro lens
Software
I’m again not sure , but I need:
- automatic picture taking with a pause for the camera to stabilize and complete its exposure
should be able to hold 5 lbs
- Because each piece of artwork is a different size, I need to be able to input a top-left coordinate and a bottom-right coordinate for limiting the area to be photographed
*because size requirements change, I need to be able to adjust the % overlap of images both vertically and horizontally. The program would then need to be able to divide the space into an appropriate overlapping virtual grid, then have the arm/camera execute the operation
*I’m expecting to have to batch stitch the images together
Troubleshooting steps already taken:
I’ve considered everything I can think of, and I’ve built manual rigs of sliders on an X/Y axis … it’s too cumbersome and more importantly, lacks uniformity across multiple pictures and setups.
Thank you so much in advance for your help!
Trying to understand the setup based on what you write, but have some questions:
- Can you put the artwork on a rotating platform, or does the setup need to work around a fixed art piece?
- Given that you’re taking many photos, are you then planning to create a 3D high resolution image?
- Do you have a budget for all this? It might take a lot of custom parts and even some machining.
Can you give a concrete (visual) example of what you might scan? For example, I’m picturing a vase on top of a pedestal with a rectangular glass enclosure at the top, and you need to place the setup next to it. Based on what you write, the setup would take photos while moving in an arc (with the art being at the center of the arc)., then move up on the z-axis and repeat.
Not the same as what you wrote, but good to be aware of products like this:
You’d need direct access to the art though (nothing blocking the scan).
Thank you so much for responding!
So the scan would be for canvas and paper-based drawings. No 3D scanning would be needed, though an extension of this project would be to create a 3D relief map of the substructure. The Scanner you posted would be a good contender for that part, but I’m concerned about the quality of the images, which need to be a minimum resolution of 300ppi to be considered archival. Excellent color range is also a must. Hence the need for a full-frame sensor DSLR or mirrorless camera.
I would have direct access and permission to implement the scans, so a physical barrier would not be an issue. I’m creating a process for scanning artwork for high-quality printing at larger-format sizes (up to or even larger than the originals scanned). I have a massive Giclee printer, so quality of the printed composite image would need to be high.
***** The Setup is a piece of artwork hung in front of the robotic arm, which would be mounted on a surface or pedestal and holding a DSLR Camera. The arm/camera would move across the artwork in an overlapping gridlike fashion to allow for accurate stitching.
I came into this need because it occurred to me that artists want to be able to print their works, but no one knows how to effectively, efficiently, and reliably scan larger pieces of artwork without laying them on a flatbed scanner, which is among many things dangerous. Also, transporting artwork in NYC is very expensive and risky, so I’d be mobile.
Does that clear it up a little? Again. thank you!!!
Not sure a robotic arm is even necessary since it’s a 2D scan. Perhaps just an X/Y positioning table mounted vertically onto which a high resolution camera is mounted. Note that you don’t require a DSRL camera to get high resolution (and a full DSLR is quite heavy). What is the largest sized canvas you foresee scanning?
