Another breakthrough in photo editing has been made by Carnegie Mellon University. They have developed software that can be of great assistance in 3D manipulations of objects in photos.
“In the real world, we’re used to handling objects – lifting them, turning them around or knocking them over… We’ve created an environment that gives you that same freedom when editing a photo.” — Natasha Kholgade, a Ph. D student in Carniege.
That’s basically the principle this innovation is rotating into. This software would allow us to see objects in photos as we do in the real world. The software can also work into paintings and human-and-objects composition. However, as this can provide picture perfect results it is not yet as flawless. The 3D photo manipulation software has certain limits like it does not offer translucency. On the other hand, the photo results are all deemed satisfactory. And it functions only with the use of already available 3D models. It means you can use it on a lot of objects, but not on everything.
How does this work? As mentioned, it uses existing 3D models’ geometry and appearance to represent the hidden parts of an object in a photo by generating points’ correspondence to align the photo with the 3D model. Yes, the software does not reveal the actual part of the object hidden from the camera, but creates a realistic angle from the given 3D model. All the same, this is a very promising tool.
Check out this video to see closely what more this 3D manipulation software can do.