How to Use the Software

There a few basic rules for what we can convert:

1. There needs to be a foreground object or person(s).  No point creating a depth image if there is no depth (e.g., with a distance landscape, where everything looks the same depth and there is nothing in the foreground).

2. There needs to be color contrast between the foreground and background.  Bluescreening is not necessary.  But if someone is camoflaged against the background, or if part of them is invisible in a shadow, a person can't see the edge of them and the software wouldn't be able to either.  

3. Frame photos to avoid the ground and ground plane when possible.

The less you bend the rules, the better the depth will be.

When the background is a multi-colored pattern or scene, it can be difficult to do edge detection around a foreground object. This video shows rapid segmentation of a foreground person using 2D-3D Photo (tm) conversion software, and fine detail like individual strands of hair being rapidly separated and modeled. Click here to see the full-size final result. No bluescreening needed.

Most image editing programs use tools like a magic wand, although these do not work well with mottled backgrounds. However, this video shows that 2D-3D Photo works very well with some pictures for picking out fine detail like fly-away hair or complex tree branches, even on a multi-colored background. The segmentation works best if there is contrast between the foreground and background, and if the image has been sharpened.

Software Overview  | Tutorial | Optimal Settings | Download | FAQs