| 摘要 |
This fully-automated 3D photo booth uses a single-or dual-lens camera or other depth sensors. Depth algorithms process the pictures taken, which can be inserted into themed templates. A novel method of coarse segmentation separates the foreground and background. The photo booth dispenses an interlaced print and snap-in photo frame, or gives the option of viewing on a mobile device or sending via email or social media. On a mobile device, the 3D viewing is with a snap-on or adhesive optical overlay if available, using novel lenticules that solve Moiré artifacts. In the absence of an optical overlay, tilt sensors are used to move the viewpoint of the 3D scene model, to look around objects in the foreground, the same as if real objects were being tilted back and forth. |
