Modeling temporal coherence for optical flow

Modeling temporal coherence for optical flow
复制标题

光流时间相干性建模

DOI:
10.1109/iccv.2011.6126359
复制
发表时间:
2011
期刊:
2011 International Conference on Computer Vision
影响因子:
--
通讯作者:
H. Zimmer
H. Zimmer
中科院分区:
--
文献类型:
--
作者:
S. Volz;A. Bruhn;L. Valgaerts;H. Zimmer

文献摘要

参考文献

被引文献

相似文献

尽管时间相干性是处理视频数据时不可否认的关键方面之一,但在最近的光流方法中,这个概念几乎没有被利用。在本文中,我们将提出一种新的参数化多帧光流计算,自然使我们能够嵌入的时间相干的空间流结构的假设,以及假设的光流是光滑的沿着运动轨迹。第一个假设是通过在多个框架上扩展空间正则化来实现的,而第二个假设是通过两个新的一阶和二阶导数平滑项来实现的。对于后者,我们调查的自适应决策方案,使本地(每像素)或全球(每序列)选择最合适的模型可能。实验表明,我们的方法相比,现有的策略施加时间的连贯性明显的优越性。此外,我们通过在广泛使用的Middlebury基准测试中获得前3名的结果来展示我们方法的最新性能。
Despite the fact that temporal coherence is undeniably one of the key aspects when processing video data, this concept has hardly been exploited in recent optical flow methods. In this paper, we will present a novel parametrization for multi-frame optical flow computation that naturally enables us to embed the assumption of a temporally coherent spatial flow structure, as well as the assumption that the optical flow is smooth along motion trajectories. While the first assumption is realized by expanding spatial regularization over multiple frames, the second assumption is imposed by two novel first- and second-order trajectorial smoothness terms. With respect to the latter, we investigate an adaptive decision scheme that makes a local (per pixel) or global (per sequence) selection of the most appropriate model possible. Experiments show the clear superiority of our approach when compared to existing strategies for imposing temporal coherence. Moreover, we demonstrate the state-of-the-art performance of our method by achieving Top 3 results at the widely used Middlebury benchmark.
DOI: --
发表时间: 2010-12
期刊: Energies
影响因子: 3.2
作者:
Deqing Sun;Erik B. Sudderth;Michael J. Black
通讯作者: Deqing Sun;Erik B. Sudderth;Michael J. Black
DOI: --
发表时间: 1995
期刊: IEEE Trans. Robotics Autom.
影响因子: --
作者:
K. Chaudhury;R. Mehrotra
通讯作者: R. Mehrotra