An active scanner measures the geometry of objects by sampling their surface with light from a controllable light source. Well known devices of this class are laser scanners, that move a laser beam over the surface. 3Digify uses a video projector for this task. In contrast to passive scanning methods, which use only camera images and no projection, active scanners are more robust and yield geometry at much higher precision.
An exemplary scanner built from a Dell M115HD video projector and two Canon EOS100D cameras.

An exemplary scanner built from a Dell M115HD video projector and two Canon EOS100D cameras.

During the last years, several tools have become available that allow to reconstruct geometry from a set of images of an object. These tools can be used for scanning an object by just taking a couple of pictures from all sides. However, successful reconstruction relies on distinct visual features of the surface and is problematic in e.g. uni-colored regions. In contrast, active scanners allow to reconstruct larger amounts of surface types at significantly higher precision.
The same object reconstructed with photogrammetry and active scanning / 3Digify. Note the significantly higher amount of details in 3Digify reconstruction.
The same object reconstructed with photogrammetry and active scanning / 3Digify. Note the significantly higher amount of details in 3Digify reconstruction.
Successful reconstruction requires the projected pattern to be clearly visible. As a rule of thumb one can say that if the pattern is visible, it is possible to reconstruct geometry. The achievable quality however strongly depends on the reflectance properties of the surface: Ideal objects are matt, opaque and do not interfere with the pattern. Paper, wood, ceramics, stones, metals and plastic work very well. Glossy surfaces usually are unproblematic too, provided that the specularity is not too high (as in chase of chrome for example). However, problematic are partially translucent materials like wax or jade. Completely transparent surfaces (glass) or mirror surfaces cannot be reconstructed. Note that these limitations do not only hold for 3Digify, but also apply for other active / passive reconstruction systems.
A minimal scanning setup consists of a single camera and a projector. However, we recommend to use an additional camera of the same type. 3Digify currently supports Canon and Nikon DSLR cameras (see supported devices). As projector, any standard video projector (HDMI, VGA, DisplayPort) may be used.
Two cameras yield significantly higher precision. It is also possible to “simulate” a two-camera setup with a single camera by moving the camera to two different positions during aquisition. However, for best handling we recommend to use a scanner with two cameras of the same kind.
The scanning precision is mainly related to the resolution of the cameras you use in your setup. Cheap high resolution cameras are available off-the-shelf, thus a scanning setup of professional quality can already be constructed at a very reasonable price using 3Digify. During the 3D reconstruction process, each pixel of the camera that perceives the light pattern will eventually be transformed into a 3D measurement. E.g. if you use a cameras with 15 megapixels (image resolution of 4752 x 3168) and your object has a width of 1 meter, you can expect a sampling density of 0,21mm (1 meter divided by 4752). If you use a camera with 18 megapixels (image resolution of 5184 x 3456) you can expect a sampling density of 0,19mm.
The same surface patch (2cm x 2cm) reconstructed with different cameras. The higher the camera resolution, the more details can be observed in the reconstructed surface.
The same surface patch (2cm x 2cm) reconstructed with different cameras. The higher the camera resolution, the more details can be observed in the reconstructed surface.
3Digify stores the reconstructed object as mesh (i.e. a set of vertices connected by faces) together with two detail maps: A texture map for surface color information and a normal map to add additional details to the surface geometry. Splitting the data into a mesh with detail maps yields a highly efficient representation (compact but of high visual quality) that may immediately be post-processed using standard CAD tools or uploaded and shared online. The mesh is provided in the PLY, OBJ or STL formats which are recognized by all CAD tools. Detail maps are provided in the BMP, JPG or PNG format. Parameters like face count and texture resolutions may be adjusted depending on your needs.
3Digify is able to calibrate the cameras automatically, which enables a very flexible scanning procedure. You simply need to place the object onto a turn target. This means you can setup the cameras and projector according to your needs (e.g. the size of the object). The object can be rotated manually in front of the scanner (or you can move the devices around the object if the object is immovable).
This depends on the speed of the cameras and on the complexity of the object. E.g. with Canon EOS 100D cameras and a turntable it takes around 20 minutes to acquire an object completely.
Yes, 3Digify produces standard CAD file formats, which can directly be processed by 3D printers.
If you have any further questions, please contact us.