Variational image motion estimation by preconditioned dual optimization

Estimating optical flows is one of the most interesting problems in computer vision, which estimates the essential information about pixel-wise displacements between two consecutive images. This work introduces an efficient dual optimization framework with accelerated preconditioners to the challenging nonsmooth optimization problem of total-variation regularized optical-flow estimation. In theory, the proposed dual optimization framework brings an elegant variational analysis to the given difficult optimization prob-lem, while presenting an efficient algorithmic scheme without directly tackling the corresponding nonsmoothness in numeric. By introducing efficient pre-conditioners with a multi-scale implementation, the proposed preconditioned dual optimization approaches achieve competitive estimation results of image motion, compared to the state-of-the-art methods. Moreover, we show that the proposed preconditioners can guarantee convergence of the implemented numerical schemes with high efficiency.