Augmented Lagrangian Digital Volume Correlation (ALDVC)

Adaptive Lagrangian Digital Volume Correlation - volumetric displacement and strain measurement based on a hybrid local-global approach

Augmented Lagrangian Digital Volume Correlation (ALDVC): volumetric displacement and strain measurement based on a hybrid local-global approach. ALDVC is a fast, parallel-computing hybrid DVC algorithm, which combines advantages of local subset method (fast computation speed, and parallel computing) and finite-element-based global method (guarantee global kinematic compatibility and decrease noise).

% ===================================

Reference paper:

For full details, and to use this code, please cite our paper:

Yang, J., Hazlett, L., Landauer, A., Franck, C. Augmented Lagrangian Digital Volume Correlation. Experimental Mechanics, 2020 (https://link.springer.com/article/10.1007/s11340-020-00607-3).

Or request full text at:

https://www.researchgate.net/publication/343676441_Augmented_Lagrangian_Digital_Volume_Correlation_ALDVC

Code manual is available at:

https://www.researchgate.net/publication/343676916_Augmented_Lagrangian_Digital_Volume_Correlation_ALDVC_Code_Manual

% ===================================

Advantages of ALDVC algorithm

[1] It’s a fast algorithm using distributed parallel computing.

[2] Global kinematic compatibility is added as a global constraint in the form of augmented Lagrangian, and solved using Alternating Direction Method of Multipliers scheme.

[3] Both displacement fields and affine deformation gradients are correlated at the same time.

[4] No need of much manual experience about choosing displacement smoothing filters.

[5] Being able to compute image sequence with multiple image frames, which is especially quite useful for measuring very large deformations.

% ===================================

% ****** ATTENTION ******

% The "x,y,z" or "1-,2-,3-" coordinates in the ALDVC code always correspond to the 1st, 2nd and 3rd indices of Matlab workspace variable. For example, p_meas(:,1) and p_meas(:,2) are the x- & y-coordinates of scattered points.

%

% This is a little different from some MATLAB image processing functions. % For example, if a 3D image has size MxNxL, in this code, we always have the image size_x=M, size_y=N, size_z=L. If you use some Matlab computer vision/image post-processing function, for example, 'imagesc3D', or 'imshow3D', or 'surf', it will reads size_x=N, size_y=M, size_z=L.

%

% Please pay attention to this difference.

% ==================================================

ALDVC example dataset

Example images download link: https://utexas.box.com/s/dha6cyk6it5h7cuxxyamq5datonw2vx4

Example results download link: https://utexas.box.com/s/nku4lkkkk95srdflhhbinurxy077jd62

Contact and support

I appreciate your comments and ratings to help me keep improving this code! Please feel free to follow this code, then you will be notified with all the important updates/corrections in the future.

% ==================================================

References

[1] regularizeNd. https://www.mathworks.com/matlabcentral/fileexchange/61436-regularizend

[2] https://www.mathworks.com/matlabcentral/fileexchange/24049-streamcolor