With a quality 3D image, you can do a lot more than with a usual 2D image.
Properties of 3D data
- Scale invariant: Since the position of the object w.r.t camera is known, its size is known. Therefore, moving closer or further from camera does not change the apparent object size, only changes the density and sampling.
- 3D data carry the geometric meaning: Bottle label, for example, can be easily "unwrapped" to its original flat form, and then processed with traditional techniques.
- Recognition of same-color object is easier: Traditional approaches with 2D have problems in segmenting connected regions of the same color. Even if there is a white item on a white background, the 3D shape is apparent and can be used to separate the object from the background.
Knowing the exact shape of the object we are looking for, we can easily match the object description with the data from the real scene. A mathematical model of the object is fit to the place in the reality. Then we are able to extract position and orientation of the matched model and provide very precise coordinates in the space and also the orientation (e.g. angle, rotation) of the object.
Various application of this include: