I'm creating a 3D map with 2D depth images in Processing. I have captured the images using saveFrame(), however I am having difficulty in converting those saved frames into 3D. Is there any website or code I could look through for help? Any help would be much appreciated.
Asked
Active
Viewed 5,062 times
1 Answers
2
Before i'm going to go in-depth into your question, i want to mention that instead of saveFrame() you can use the standard dfx library to export 3d models instead of 2d images using Processing if you simply want to store a scene: https://processing.org/reference/libraries/dxf/
Now back to your question. First of all what are depth images? Are those simply saveFrames from a 3D Scene in Processing (P3D) or are these special images, because depth is quite a general term. If they are 3D Scenes and you would know the coordinates of the camera and their viewangle the task gets quite easier, but it is technically impossible to create a 3D object using only 2D images without XRay. Imagine looking at a fork. Your eyes are making 2 pictures of that fork, however you have no idea what might be inscribed on the back of that fork. No matter how many pictures you might have of your 3D scene, you won't be able to convert it into 3D perfectly. That said, this is indeed a general problem in computer science and there are various methods to solve this. Wikipedia has an article on it: http://en.wikipedia.org/wiki/3D_reconstruction_from_multiple_images http://en.wikipedia.org/wiki/2D_to_3D_conversion
Here are a few stackoverflow topics which might help you get started:
3d model construction using multiple images from multiple points (kinect)
