Abstract | ||
---|---|---|
•We propose a statistical appearance model without one-to-one correspondences.•Model generation and adaption are expressed by a closed-form optimization criterion.•No prior correspondence determination is necessary.•Model parameters and probabilistic correspondences are optimized iteratively.•The model shows high robustness for problems with missing structures/occlusions. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1016/j.media.2017.02.004 | Medical Image Analysis |
Keywords | Field | DocType |
Active appearance model,Statistical shape model,Probabilistic correspondences,Model-based segmentation | Active shape model,Computer vision,Feature vector,Pattern recognition,Segmentation,Image processing,Active appearance model,Preprocessor,Sparse image,Artificial intelligence,Probabilistic logic,Mathematics | Journal |
Volume | ISSN | Citations |
37 | 1361-8415 | 3 |
PageRank | References | Authors |
0.43 | 16 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Julia Krüger | 1 | 16 | 7.63 |
Jan Ehrhardt | 2 | 387 | 54.33 |
Heinz Handels | 3 | 1527 | 239.84 |