Article from:  Arduino-powered desktop 3D scanner 

Think it’s possible to build a desktop 3D scanner with infinite resolution in budget? Instructables guru Whitney Potter has done just that, using a Nikon DSLR and an Arduino-powered stepper motor. To capture its high-res scans, Potter’s systems uses photogrammetry—a method of constructing 3D images from a set of partially overlapping 2D photographs. The impressive results have to be seen to be believed.

Naturally, the main caveat of the “3D scanner” claim is its assumption of prior DSLR ownership. Builders of the budget scanner will need their own high-quality digital camera—unlikely to be found within that strict budget! Once that small condition is met, however, the process becomes extremely wallet-friendly. The principle of Potter’s system is simple: Why buy a dedicated 3D scanner when, with the right manipulation, your high-quality digital camera is capable of producing superior results, and on subjects of all sizes?

The main obstacles faced during photogrammetry are photograph quality and spacing. Each photo must be perfectly focused and exposed, and must be spaced around the object to capture its every feature. There must be sufficient overlap between photos so the rendering software knows where each shot belongs. Getting this spacing right can be achieved with some practice on large subjects, but is incredibly difficult with small objects. This is where Potter’s Arduino-powered stepper motor comes in.

The DIY 3D scanner uses a stepper motor controlled by an Arduino board to turn the scanned object by a fixed amount, before an LED triggers the camera’s wireless sensor, setting off the shutter. This process is repeated until photos have been collected from all angles, allowing one’s photogrammetry software to construct an accurate and high-resolution 3D image of the subject. There are just five key ingredients to the budget 3D scanner: Arduino, stepper motor driver, stepper motor, IR LED and LCD shield.

Although Potter’s Arduino sketch has been configured to work with Nikon DSLRs, owners of other brands can simply alter a section of the code to suit their own model. Potter has used Sebastian Setz’s Multi-Camera IR Control library, so the scanner will work with virtually any brand of camera with an IR remote. When the code is in place, the subject needs to be well-lit so the camera can capture all of its precise details. The captured set of photos can then be processed using photogrammetry software. Potter recommends Agisoft Photoscan, Autodesk Memento or Autodesk 123D Catch (budget option).

Potter’s handy guide demonstrates an innovative method for 3D scanning on a budget—using resources at most people’s disposal.
Could DIY photogrammetry be the future of desktop 3D scanning? Watch out, 3D scanner manufacturers!