5.5. Tensor Voting in ND
The framework can easily be generalized to higher dimensions to tackle perceptual organization problems in ND spaces. At first glance, it might not be apparent how computer vision problems can be formulated in domains with more than three dimensions, but some computer vision problems are very naturally represented in such domains. For instance, in Section 5.6, we show how optical flow can be formulated as the organization of tokens in smooth layers in a 4D space, where the four dimensions are the image coordinates and the horizontal and vertical velocity.
Both the representation and voting can be generalized to N dimensions along the lines of the previous section, where the generalization from 2D to 3D was shown. The ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access