Feature Extraction

In principle, the whole orientation histogram can be used to characterize the texture of 3D data. However, this is a rather cumbersome representation and ideally one would like to use only a few features to describe its shape. Perhaps the grosser characteristic one may observe in the orientation histogram is its symmetry: A spherically symmetric histogram implies a totally isotropic volume, whereas any deviation from that is a cue to the anisotropic structure of the underlying material. In what follows, therefore, we use the orientation histogram in order to characterize the anisotropy of the data.

Was this article helpful?

0 0

Post a comment