A 117 Line 2D Digital Image Correlation Code Written in MATLAB
Round 1
Reviewer 1 Report
The paper “A 117 Line 2D Digital Image Correlation Code Written in MATLAB” presents a 2D DIC algorithm developed by exploiting the common workflow in digital image correlation. The paper firstly explains the DIC theory, and then explains how the theory is implemented in a Matlab software.
In reviewer’s opinion, the topic of the paper is of interest for the scientific community, since DIC is widespread in many research fields, and this paper provides a clear and concise overview of the theory behind the method. The overall quality of the paper is good (organization, language, notation) and it is scientifically sound. In Reviewer’s opinion, the paper can be useful for researchers approaching DIC methods willing to improve the algorithm itself, and can be considered a didactic resource.
Some issues were found in the paper, which are listed below (the order corresponds to the position in the paper):
- Introduction
The Authors provide many references about the research fields where DIC is used. Anyway, even if they talk about issues related to vibrations, they do not add any references about DIC application to the vibration measurements. Some recent papers deals with the topic, such as:
- a) S. Barone, P. Neri, A. Paoli, A.V. Razionale, Low-frame-rate single camera system for 3D full-field high-frequency vibration measurements, Mechanical systems and signal processing, 2019, vol.123, pp.143-152
- b) A.J. Molina-Viedma, L. Felipe-Sese, E. Lopez-Alba, F.A. Diaz, 3D mode shapes characterisation using phase-based motion magnification in large structures using stereoscopic DIC, Mechanical systems and signal processing, 2018, vol.108, pp.140–155
- Figure 3
The Reviewer suggests to add the SF order corresponding to each sub-image directly on top of the image.
- Line 443
Formatting issue (not-needed return after “defined in”).
- The presented algorithm allows to set different DIC parameters. While the effect of each parameter is explained in the paper, the Authors should add some discussion about the subset shape: which are the cases that requires a square subset or a circular subset?
- Validation
The Authors present numerical results in terms of displacements error by comparing different algorithms. While the presented tables are sure useful to quantitatively compare the performances, some displacement maps could be of interest to also obtain a visual comparison. Also, the effect of contrast and noise on the performance of different algorithm is discussed, nut the Authors do not discuss the effect of displacement amplitude: which method performs better with smaller displacement?
- Line 687
Typo: “This is repeated until the all the subsets have been correlated” remove the first “the”.
- Validation
In Reviewer’s understanding, the validation was performed by using synthetic images. This is surely useful to separate different effects and to assess the algorithm robustness. Nevertheless, testing the algorithms on actual images, corresponding to some physical experiment, would be of interest for the reader.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The article named "A 117 Line 2D Digital Image Correlation Code Written in MATLAB" published by the authors Devan Atkinson and Thorsten Becker has a logical form and contains an appropriate and high-quality summary of the 2D digital image correlation theory with its mathematical background. Moreover, it includes a description of the unique 117 line code created in Matlab serving for the realisation and evaluation of the measurement using single-camera digital image correlation system. Several analysis, by which the subset size was varying, were performed to validate the results obtained by the proposed code. The quantified errors prove the good correspondance with Ncorr and LaVision's DaVis software.
It is an excellent scientific paper and it can be published if the following minnor revisions will be adressed:
1) I cannot find any information about the calibration process. Can you please add the information, how does the calibration process work? What kind of calibration target is required? Are there any specific demands on the shape of the calibration target?
2) Can the authors establish the usability of the proposed code on a real measurement (e.g. strain analysis of the flat specimen loaded by simple uniaxial loading) and make a qualitative comparison of the results with a numerical analysis?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
Although I did (could) not retrace all coherences and formulas the paper appears to me as a good and comprehensive explanation of DIC and it's implementation into a MATLAB framework.
Recommendation: Expand the chapter 4 by some real-world images (maybe from images used in [37]) and the presentation of MATLAB code by a short example script calling the functions in Appendix A. This would make the paper even better and attractive not only for people interested in mathematics.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
