Low complexity dense motion estimation using phase correlation

Authors:
V. Argyriou;T. Vlachos
Affiliations:
School of Computing, Information Technology and Engineering, University of East London, UK;Department of Audiovisual Arts, Ionian University, Greece
Venue:
DSP'09 Proceedings of the 16th international conference on Digital Signal Processing
Year:
2009

Citing 12
Cited 0

Computing optical flow across multiple scales: an adaptive coarse-to-fine strategy

International Journal of Computer Vision
Performance of optical flow techniques

International Journal of Computer Vision
Robust computation of optical flow in a multi-scale differential framework

International Journal of Computer Vision
The robust estimation of multiple motions: parametric and piecewise-smooth flow fields

Computer Vision and Image Understanding
Robot Vision

Robot Vision
Multimodal Estimation of Discontinuous Optical Flow using Markov Random Fields

IEEE Transactions on Pattern Analysis and Machine Intelligence
Lucas-Kanade 20 Years On: A Unifying Framework

International Journal of Computer Vision
The Dense Estimation of Motion and Appearance in Layers

CVPRW '04 Proceedings of the 2004 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'04) Volume 11 - Volume 11
Image alignment and stitching: a tutorial

Foundations and Trends® in Computer Graphics and Vision
Normalized averaging using adaptive applicability functions with applications in image reconstruction from sparsely and randomly sampled data

SCIA'03 Proceedings of the 13th Scandinavian conference on Image analysis
Robust optical flow estimation based on a sparse motion trajectory set

IEEE Transactions on Image Processing
Dense motion estimation using regularization constraints on local parametric models

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose a low-complexity dense motion estimation scheme particularly attractive for real-time video applications. Our scheme is based on overlapped block-based motion estimation using phase correlation at critical pixel locations. These form an irregularly sampled grid capturing salient motion features of a scene. The dense vector field is obtained by applying normalized convolution on the irregular grid. Our experiments show that our scheme provides reliable sub-pixel accuracy motion vectors corresponding to actual scene motion, outperforms differential and phase-based methods and yields comparable performance to more complex and time consuming robust motion estimation techniques.