Curve-Skeleton Properties, Applications, and Algorithms
IEEE Transactions on Visualization and Computer Graphics
Learning parameter tuning for object extraction
ACCV'06 Proceedings of the 7th Asian conference on Computer Vision - Volume Part I
| Hi-index | 0.00 |
We present a robust method for extracting 3D centerlines from volumetric datasets. We start from a 2D skeletonization method to locate voxels centered with respect to three orthogonal slicing directions. Next, we introduce a new detection criterion to extract the centerline voxels from the above skeletons, followed by a thinning, reconnection, and a ranking step. Overall, the proposed method produces centerlines that are object-centered, connected, one voxel thick, robust with respect to object noisiness, handles arbitrary object topologies, comes with a simple pruning threshold, and is fast to compute. We compare our results with two other methods on a variety of real-world datasets.