Quaternion Atomic Phase Magnification for 2D and 3D Motion

The phase computation is used in many applications of signal and image processing such as edge, line, symmetry detection, image analysis and recently to show small motion or color changes. Eulerian video processing allows to show small color changes and motions invisibles to the naked eye. Eulerian motion magnification, was modify using the phase and the Riesz pyramid in order to reduce the noise amplification. In this work we introduced a novel phase magnification approach based on a infinitely differentiable unique compact supported window, an atomic function. Another contribution of our work was apply our quaternion phase and Eulerian magnification to 3D motion, which has not been reported so far, we propose to use simple projections of conformal algebra in order to do the magnification at any plane of the 3D space. We compare the phase (Riesz Pyramids), with the atomic Riesz pyramid, in our comparation our atomic pyramid Riesz preserves more information than the phase Riesz approximation, besides our approach is slower by 2.5 times. Finally, we compare method and the linear magnification (Laplace Pyramid) in terms of the Signal to Noise Ratio (SNR). We found that the phase magnifications has better response to noise than the linear method.

[1]  Stephen J. Sangwine,et al.  Hypercomplex Fourier Transforms of Color Images , 2001, IEEE Transactions on Image Processing.

[2]  Eduardo Bayro-Corrochano,et al.  Quaternion Atomic Function for Image Processing , 2011, Guide to Geometric Algebra in Practice.

[3]  Stephen J. Sangwine,et al.  Quaternion and Clifford Fourier Transforms and Wavelets , 2013 .

[4]  Michael Rubinstein,et al.  Analysis and visualization of temporal variations in video , 2014 .

[5]  Eduardo Ulises Moya-Sánchez,et al.  Hilbert and Riesz Transforms Using Atomic Function for Quaternionic Phase Computation , 2013, Advances in Applied Clifford Algebras.

[6]  Peter Kovesi,et al.  Invariant measures of image features from phase information , 1996 .

[7]  Michael Felsberg,et al.  Low-level image processing with the structure multivector , 2002 .

[8]  John Weiss The Hilbert Transform of Wavelets are Wavelets � , 2002 .

[9]  Michael Felsberg,et al.  The monogenic signal , 2001, IEEE Trans. Signal Process..

[10]  Eckhard Hitzer,et al.  Quaternion Fourier Transform on Quaternion Fields and Generalizations , 2007, ArXiv.

[11]  Frédo Durand,et al.  Quaternionic Representation of the Riesz Pyramid for Video Magnification , 2014 .

[12]  Frédo Durand,et al.  Visual vibrometry: Estimating material properties from small motions in video , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Thomas Bülow,et al.  Hypercomplex spectral signal representations for the processing and analysis of images , 1999 .

[14]  Eduardo Bayro-Corrochano,et al.  The Theory and Use of the Quaternion Wavelet Transform , 2005, Journal of Mathematical Imaging and Vision.

[15]  Frédo Durand,et al.  Riesz pyramids for fast phase-based video magnification , 2014, 2014 IEEE International Conference on Computational Photography (ICCP).

[16]  Bill Casselman,et al.  Clifford algebras and spinors , 2017 .

[17]  Anne Sedlazeck,et al.  Local Feature Detection by Higher Order Riesz Transforms on Images , 2008 .