Particle reconstruction of volumetric particle image velocimetry with the strategy of machine learning

Three-dimensional particle reconstruction with limited two-dimensional projections is an under-determined inverse problem that the exact solution is often difficult to be obtained. In general, approximate solutions can be obtained by iterative optimization methods. In the current work, a practical particle reconstruction method based on a convolutional neural network (CNN) with geometry-informed features is proposed. The proposed technique can refine the particle reconstruction from a very coarse initial guess of particle distribution that is generated by any traditional algebraic reconstruction technique (ART) based methods. Compared with available ART-based algorithms, the novel technique makes significant improvements in terms of reconstruction quality, robustness to noise, and at least an order of magnitude faster in the offline stage.

[1]  T. Tang,et al.  Effect of water injection on the cavitation control:experiments on a NACA66 (MOD) hydrofoil , 2020, Acta Mechanica Sinica.

[2]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[3]  Nicholas A. Worth,et al.  Acceleration of Tomo-PIV by estimating the initial volume intensity distribution , 2008 .

[4]  G. N. Minerbo,et al.  MENT: A maximum entropy algorithm for reconstructing a source from projection data , 1979 .

[5]  Michael Unser,et al.  Convolutional Neural Networks for Inverse Problems in Imaging: A Review , 2017, IEEE Signal Processing Magazine.

[6]  B. Wieneke.  Iterative reconstruction of volumetric particle distribution , 2013 .

[7]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[8]  Christoph Schnörr,et al.  Recasting Tomo-PIV Reconstruction as Constrained and L1-Regularized Non-Linear Least Squares Problem , 2010 .

[9]  R H Huesman,et al.  The effects of a finite number of projection angles and finite lateral sampling of projections on the propagation of statistical errors in transverse section reconstruction. , 1977, Physics in medicine and biology.

[10]  Olivier Alata,et al.  Tomographic Reconstruction of 3D Objects Using Marked Point Process Framework , 2018, Journal of Mathematical Imaging and Vision.

[11]  Wei Xu,et al.  CNN-RNN: A Unified Framework for Multi-label Image Classification , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Karthik Duraisamy,et al.  Prediction of aerodynamic flow fields using convolutional neural networks , 2019, Computational Mechanics.

[13]  Qi Gao,et al.  Review on development of volumetric particle image velocimetry , 2013 .

[14]  Chao Xu,et al.  Particle Image Velocimetry Based on a Deep Learning Motion Estimator , 2020, IEEE Transactions on Instrumentation and Measurement.

[15]  Aliza Abraham,et al.  Snow-powered research on utility-scale wind turbine flows , 2019, Acta Mechanica Sinica.

[16]  Neena Aloysius,et al.  A review on deep convolutional neural networks , 2017, 2017 International Conference on Communication and Signal Processing (ICCSP).

[17]  Yoshua Bengio,et al.  Convolutional networks for images, speech, and time series , 1998 .

[18]  Jinjun Wang,et al.  Dual-basis reconstruction techniques for tomographic PIV , 2015 .

[19]  Wei Li,et al.  Convolutional Neural Networks for Steady Flow Approximation , 2016, KDD.

[20]  Bernhard Wieneke,et al.  Tomographic particle image velocimetry , 2006 .

[21]  Bernhard Wieneke,et al.  Volume self-calibration for 3D particle image velocimetry , 2008 .

[22]  R B Guenther,et al.  Reconstruction of objects from radiographs and the location of brain tumors. , 1974, Proceedings of the National Academy of Sciences of the United States of America.

[23]  Jinjun Wang,et al.  Intensity-enhanced MART for tomographic PIV , 2016 .

[24]  Stefano Discetti,et al.  Spatial filtering improved tomographic PIV , 2013 .

[25]  G. Elsinga,et al.  Ghost hunting—an assessment of ghost particle detection and removal methods for tomographic-PIV , 2014 .

[26]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[27]  Kookjin Lee,et al.  Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders , 2018, J. Comput. Phys..

[28]  Ivan Marusic,et al.  Enhancing Tomo-PIV reconstruction quality by reducing ghost particles , 2013 .

[29]  Hanspeter Pfister,et al.  Parallel Separable 3D Convolution for Video and Volumetric Data Understanding , 2018, BMVC.

[30]  Seyed-Ahmad Ahmadi,et al.  V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[31]  Julio Soria,et al.  An efficient simultaneous reconstruction technique for tomographic particle image velocimetry , 2009 .

[32]  Jiaming Liang,et al.  Filtering enhanced tomographic PIV reconstruction based on deep neural networks , 2020, IET Cyber-Systems and Robotics.

[33]  Fulvio Scarano,et al.  Tomographic PIV: principles and practice , 2012 .

[34]  Fulvio Scarano,et al.  An efficient and accurate approach to MTE-MART for time-resolved tomographic PIV , 2015 .

[35]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[36]  A. Schröder,et al.  Shake-The-Box: Lagrangian particle tracking at high particle image densities , 2016, Experiments in Fluids.

[37]  Andreas Schröder,et al.  Shake The Box’ - a 4D PTV algorithm: Accurate and ghostless reconstruction of Lagrangian tracks in densely seeded flows , 2014 .

[38]  Kurt Keutzer,et al.  Spatially Parallel Convolutions , 2018, ICLR.

[39]  Alexandra H. Techet,et al.  Fast volume reconstruction for 3D PIV , 2017 .

[40]  Xiaolin Hu,et al.  Recurrent convolutional neural network for object recognition , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[41]  Zenghui Wang,et al.  Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review , 2017, Neural Computation.

[42]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[43]  Chao Xu,et al.  Dense motion estimation of particle images via a convolutional neural network , 2019, Experiments in Fluids.

[44]  Liu Jun,et al.  A review of object detection based on convolutional neural network , 2017, 2017 36th Chinese Control Conference (CCC).

[45]  Kees Joost Batenburg,et al.  Motion tracking-enhanced MART for tomographic PIV , 2010 .