An Unbiased Second-Order Prior for High-Accuracy Motion Estimation
Proceedings of the 30th DAGM symposium on Pattern Recognition
Convex Hodge Decomposition of Image Flows
Proceedings of the 30th DAGM symposium on Pattern Recognition
Convex Hodge Decomposition and Regularization of Image Flows
Journal of Mathematical Imaging and Vision
Total-Variation Based Piecewise Affine Regularization
SSVM '09 Proceedings of the Second International Conference on Scale Space and Variational Methods in Computer Vision
On a Decomposition Model for Optical Flow
EMMCVPR '09 Proceedings of the 7th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition
Variational method for super-resolution optical flow
Signal Processing
Wavelet-Based fluid motion estimation
SSVM'11 Proceedings of the Third international conference on Scale Space and Variational Methods in Computer Vision
| Hi-index | 0.00 |
We study the estimation and decomposition of optical flows from highly nonrigid motions. To this end, recent methods from image decomposition into structural and textural parts are combined with variational optical flow estimation. The approaches we suggest amount to minimizing discrete convex functionals using second-order cone programming. Higher-order regularization is necessary in order to accurately recover important flow structure like vortices, and to incorporate key physical properties such as vanishing divergence. For proper discretization, we apply the finite mimetic difference method, which preserves the identities fulfilled by the continuous differential operators. Numerical examples demonstrate the feasibility of the complex approaches.