Postprocessing of Optical Flows Via Surface Measures and Motion Inpainting

Kondermann, Claudia; Kondermann, Daniel; Garbe, Christoph S.

doi:10.1007/978-3-540-69321-5_36

Cited by 10 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Anandan et al [1] evaluate the dependence of the sum of square differences (SSD) criterion on the displacement and define a confidence criterion as a function of principal curvatures of the SSD surface and the SSD value at minimum. In a unifying way, the 'surface measures' of Kondermann et al [20] use principal curvatures to analyze the intrinsic dimensions [43] of image invariance functions based on brightness, SSD, gradient, and Hessian.…”

Section: Related Workmentioning

confidence: 99%

“…Unreliable locations (outliers) are found by thresholding the confidence measure, where a suitable threshold can be determined statistically. Then an improved optical flow can be found by motion inpainting [20], or, as we have done here, by rerunning the optical flow computation given by Eqs. (9)- (11) with b i = 1 for pixels to be kept and b i = 0 for pixels to be ignored.…”

Section: Reliable Pixel Selectionmentioning

confidence: 99%

See 1 more Smart Citation

Bootstrap optical flow confidence and uncertainty measure

Kybic

Nieuwenhuis

2011

Computer Vision and Image Understanding

View full text Add to dashboard Cite

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues.Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. a b s t r a c tWe address the problem of estimating the uncertainty of optical flow algorithm results. Our method estimates the error magnitude at all points in the image. It can be used as a confidence measure. It is based on bootstrap resampling, which is a computational statistical inference technique based on repeating the optical flow calculation several times for different randomly chosen subsets of pixel contributions. As few as ten repetitions are enough to obtain useful estimates of geometrical and angular errors. For demonstration, we use the combined local-global optical flow method (CLG) which generalizes both LucasKanade and Horn-Schunck type methods. However, the bootstrap method is very general and can be applied to almost any optical flow algorithm that can be formulated as a pixel-based minimization problem. We show experimentally on synthetic as well as real video sequences with known ground truth that the bootstrap method performs better than all other confidence measures tested.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Reliable Pixel Selectionmentioning

confidence: 99%

Bootstrap optical flow confidence and uncertainty measure

Kybic

Nieuwenhuis

2011

Computer Vision and Image Understanding

View full text Add to dashboard Cite

show abstract

“…This will naturally lead to sparse motion fields. For applications, in which dense fields are required, a subsequent parameter field interpolation can be conducted [18]. The required regularizer can be any of those presented by [26].…”

Section: Local Spatiotemporal Approachmentioning

confidence: 99%

Spatiotemporal Image Analysis for Fluid Flow Measurements

Garbe

Kondermann

Jehle

et al. 2009

Notes on Numerical Fluid Mechanics and Multidisciplinary Design

Self Cite

View full text Add to dashboard Cite

Abstract. In this chapter, a framework will be presented for measuring and modeling transport processes using novel visualization techniques and extended optical flow techniques for digital image sequence analysis. In this way, parameters besides the 2-D xy velocity components can be extracted concurrently from the acquired 2-D image sequences, such as wall shear rates and momentum transport close to boundaries, diffusion coefficients, and depth z in addition to the z velocity components. Depending on the application, particularly the temporal regularization can be enhanced, leading to stabilization of results and reduction of spatial regularization. This is frequently of high importance for flows close to boundaries. Results from applications will be presented from the fields of environmental and life sciences as well as from engineering.

show abstract

“…In order to inpaint the flow in the occlusion areas, Matsushita et al [17] and Strobel et al [26] extended the Telea-inpainting method [27] to optical flow, i.e., they assume that the motion variation is locally small and propagate the optical flow according to a weighting function which depends on the Euclidean distance and the color difference among the interpolated pixel and its neighbours. Kondermann et al [13] proposed a postprocess of the optical flow in order to improve it: the optical flow is retained at points where it is reliable and is then densified by minimizing the L 2 norm of the spatio-temporal gradient of the flow. Berkels et al [6] proposed to recover the optical flow in non-reliable regions using a TV-type anisotropic functional and a rotation-invariant regularizer was proposed by Palomares et al [21].…”

Section: Introductionmentioning

confidence: 99%

Motion Inpainting by an Image-Based Geodesic AMLE Method

Oliver

Raad

Ballester

et al. 2018

2018 25th IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

This work presents an automatic method for optical flow inpainting. Given a video, each frame domain is endowed with a Riemannian metric based on the video pixel values. The missing optical flow is recovered by solving the Absolutely Minimizing Lipschitz Extension (AMLE) partial differential equation on the Riemannian manifold. An efficient numerical algorithm is proposed using eikonal operators for nonlinear elliptic partial differential equations on a finite graph. The choice of the metric is discussed and the method is applied to optical flow inpainting and sparse-to-dense optical flow estimation, achieving top-tier performance in terms of End-Point-Error (EPE).

show abstract

Postprocessing of Optical Flows Via Surface Measures and Motion Inpainting

Cited by 10 publications

References 17 publications

Bootstrap optical flow confidence and uncertainty measure

Bootstrap optical flow confidence and uncertainty measure

Spatiotemporal Image Analysis for Fluid Flow Measurements

Motion Inpainting by an Image-Based Geodesic AMLE Method

Contact Info

Product

Resources

About