The app works by incorporating four different photographic samples with different light sources and mathematically builds a 3-D composite. On the iPhone the light source is the screen and the capture device is the front camera. When it starts scanning, a pattern with a white area on a black background appears on the screen. It switches through four different patterns, each with the white area in different places. These would be the four different light sources.
Obviously, with light sources so close to one-another, the object to be scanned also needs to be fairly small and close to the screen for any meaningful depth to be captured. The developers recommend no more than 20 centimeters. The object to be scanned can be as large as a human head.
The capture is done fairly quickly. It takes less than two seconds to scan, and just as quickly, the composite is built almost instantly and is displayed immediately. The result is a bona-fide 3-D object built with 3-D vectors with a skin overlaid on top.
You can remove the skin so that you can see a wireframe representation. As you rotate the image by swiping left and right, it is clearly verifiable that the points have x, y, and z coordinates, making for a true, 3-D digital image. It is truly amazing that this technology actually works.
From a practical perspective though, this app is not going to replace any dedicated 3-D scanner. The resolution is quite rough and provides detail that is equivalent to that of most 3-D games. Its accuracy is also a bit rough, which you’ll mostly see in slight distortion of proportions. Also, it obviously does not do a full 360-degree capture, as it can only capture one side of an object with maybe five degrees of parallax.
This app would be much more useful if the scanned objects could be exported for use in 3-D modeling software. In this debut version it does not have that function, and it is currently most useful as an educational tool or a tech demo.
Trimensional is available for $0.99.
