User Tools

Site Tools


faq:en_possibilities

← Back to main FAQ page

What is possible with DAVID

What can I scan?

Generally, all visible matte surfaces that do not move for at least a few seconds (scanning time).

DAVID can create 360° scans by combining single scans (views from different sides). The scans should overlap to make correct alignment easy.

DAVID captures the 3D surface as a triangle net (point cloud plus connecting triangles) at correct scale.

Textures can be recorded too (with the same camera, one more click) so you can get a colored model.

To get some impressions, please take a look at our User Gallery.

Sizes

DAVID is mainly used for objects between 30 mm and 500 mm. However the whole setup is scalable, and other sizes are possible if you adapt your hardware. Scan sizes larger than 2 meter will require special setups.

You will need to calibrate the scanner for the given size, which requires a 90° calibration corner in about the same size as the object.

It is possible to scan large objects by scanning smaller regions separately and combining them later.

Extremely small

Some examples from our User Forum: Micro Scans Dollar coin Insect Fingerprints

Extremely large

Some examples from our User Forum: Full size car Back of a car Another car Dinosaur skeleton

Surfaces

The surface must not be too shiny. Highly reflective, translucent or very dark surfaces must be covered with coating spray.

Resolution, Accuracy

Depends very much on your hardware. Basically the resolution is one 3D vertex per camera pixel. So a 1 megapixel camera will give you up to 1 million 3d points / vertices. Example: Object size 100 mm, 1.3 megapixel camera –> resolution around 0.1 mm.

The expected accuracy can be even better than the resolution because we use subpixel accuracy algorithms. However the hardware and calibration must be precise enough…

There have been some investigations about absolute accuracy in our forum here and here.

I want to scan something special: ...

... human faces

No problem. We recommend to use Structured Light (not laser scanning) because it's faster and not dangerous to the eyes.

Some examples: Johanna Walter BLinus Greg

... human bodies

We recommend to use Structured Light (not laser scanning) because it's faster. Scanning regions (like face or back) is easy. Scanning the whole body requires either a large setup or combining several scans.

You may need more than one scan (e.g. to get a 360° model, or the face including both ears). In that case the person should not move between scan, the body should be as rigid as possible. Otherwise the single scans won't fit to each other.

Some body scans in our forum: Linus Ben Dennis Gunter

... metal parts

If the parts are too shiny (or too dark), they won't reflect the projector/laser light correctly. For this case we have whiting and coating sprays to cover the object with.

... jewelry

Rather difficult. With standard hardware you will get a resolution of 0.1 mm. Shiny objects need to be covered with coating spray, which can sometimes be difficult to wash out of small grooves. Deep undercuts are difficult to obtain.

... cars

You should use Structured Light (not laser scanning) because it does not require calibration panels as background during scanning. You'll need a very large setup, or (better) scan separate small regions (below 1 meter * 1 meter), overlapping, and combine them later in the software. Shiny surfaces have to be covered with coating spray.

Some results in our forum: Full size Jaguar Back of a Toyota Volkswagen

... undercuts, deep holes

Possible with limitations: DAVID uses triangulation, that means that camera and light source (laser or projector) must have a certain distance between them, they must look from different directions (at least 10° between them). You can only obtain surface regions which can be seen by both, camera and light source, at the same time.

A simple comparison: You can scan only surface regions that you can see with BOTH your eyes at the same time.

... outdoors

Possible. Bright daylight or direct sunlight must be avoided. On the scanned surface, the light source (projector or laser) must be significantly brighter than the environment light. Also you'll need power supply for computer and laser/projector.

Here is an example (please imagine a Laptop computer ;-): Outdoor scanning.

What can I do with the scan data?

Can the data be used for 3D printing or milling?

Yes! Of course DAVID does not create a machine program that drives your printer/CNC, but usually these machines come with a software that can import 3D triangle meshes.
Usually they require a “water-tight” 3D model (a model which has no holes, so that “inside” and “outside” are clearly defined). DAVID's single scans are not closed, but the last processing step in DAVID, the Fusion, can create water-tight models.
You can download some example scan results to see if you can print them.

Can the data be used for measurements/comparison/analysis?

Yes! At this time the DAVID software does not include any measurement or analysis functions, but you can export in standard 3D file formats (OBJ, STL, PLY) and use 3rd party software for analysis. The 3D data are at absolute scale (usually in mm).

Which software can I use for 3D measurements?

Meshlab is free and offers e.g. point-to-point distance measurements. Other powerful 3D software are Rhino3d, Geomagic, …

Can I export point clouds?

DAVID exports to the OBJ file format. It is an ASCII text file that contains a 3D point cloud (plus triangles, surface normals, texture coordinates). If you are interested only in the point cloud, read all lines that start with “v”. Each line stores the X, Y, Z coordinates of one 3D point. The reference coordinate system is the center of the calibration corner, with the Y axis pointing upward.

Can I import the results in my 3D software XYZ?

Please look for an import function for OBJ, STL or PLY in your software. For some programs, an import plugin (with our without costs) may be necessary.
Alternatively you may be able to find a conversion tool to convert OBJ or STL into the format that you need.
You can download some example scan results (OBJ and STL) to see if you can import them in your software.

faq/en_possibilities.txt · Last modified: 2014/06/06 18:38 (external edit)