Fast piecewise-affine motion estimation without segmentation
Current algorithmic approaches for piecewise affine motion estimation are based on alternating motion segmentation and estimation. We propose a new method to estimate piecewise affine motion fields directly without intermediate segmentation. To this end, we reformulate the problem by imposing piecew...
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Dokumenttyp: | Article (Journal) |
| Sprache: | Englisch |
| Veröffentlicht: |
23 July 2018
|
| In: |
IEEE transactions on image processing
Year: 2018, Jahrgang: 27, Heft: 11, Pages: 5612-5624 |
| ISSN: | 1941-0042 |
| DOI: | 10.1109/TIP.2018.2856399 |
| Online-Zugang: | Verlag, lizenzpflichtig, Volltext: https://doi.org/10.1109/TIP.2018.2856399 Verlag, lizenzpflichtig, Volltext: https://ieeexplore.ieee.org/document/8417969 
|
| Verfasserangaben: | Denis Fortun, Martin Storath, Dennis Rickert, Andreas Weinmann, and Michael Unser, Fellow, IEEE |
| Zusammenfassung: | Current algorithmic approaches for piecewise affine motion estimation are based on alternating motion segmentation and estimation. We propose a new method to estimate piecewise affine motion fields directly without intermediate segmentation. To this end, we reformulate the problem by imposing piecewise constancy of the parameter field, and derive a specific proximal splitting optimization scheme. A key component of our framework is an efficient 1D piecewise-affine estimator for vector-valued signals. The first advantage of our approach over segmentation-based methods is its absence of initialization. The second advantage is its lower computational cost, which is independent of the complexity of the motion field. In addition to these features, we demonstrate competitive accuracy with other piecewise-parametric methods on standard evaluation benchmarks. Our new regularization scheme also outperforms the more standard use of total variation and total generalized variation. |
|---|---|
| Beschreibung: | Gesehen am 08.04.2020 |
| Beschreibung: | Online Resource |
| ISSN: | 1941-0042 |
| DOI: | 10.1109/TIP.2018.2856399 |