Advancements in Roundness Measurement Parts for Industrial Automation Using Internet of Things Architecture-Based Computer Vision and Image Processing Techniques

Questions

Comments

Previous Line No.

New line No.

Authors propose a roundness measurement technique for industrial automation by leveraging an IoT architecture, computer vision and image processing. From my point of view, like any proposal for a new measurement system, this must be accompanied, imperatively, by an estimation of the uncertainties inherent in the inspection method. Also, it is well known that the algorithmic error must be evaluated by probing the influence of the algorithm's hyper parameters. This aspect is completely overlooked in research.

MAJOR: What will be the uncertainty of the proposed method?

Thanks for your fruitful comment.  We base our decisions on an understanding of the entire measurement process, which includes data acquisition, image processing using computer vision, and physical measurements in the case of CMM. Nevertheless, potential sources of uncertainty within the measurement process include lighting, calibration errors, image noise in CV, and machine repeatability in CMM when environmental conditions are taken into account. Furthermore, statistical analysis is used to demonstrate error propagation using SPSS software. In addition, by comparing the data from the two systems and explaining how the factors in each system contribute to measurement uncertainty.

----

------

I noted many typo mistakes (e.g., all mathematical symbols must be done with the same math editor, e.g., xi, xo (line 163), sometimes equations are BOLD? Eq. 5 to eq 9., etc.); Mathematical symbols need to be standardised in the whole manuscript.

Thank you for your fruitful comment. We had changed the typos and used the proper one of equations standard. (highlighted in red color)

163

187,

483-502

Abstract: terms used ‘metric and pixel units’… I’m not sure that is appropriate.

We appreciate your great comment. We had changed to appropriate way of writing abstract and added this sentence “It is calibrated to accommodate various units of measurement and has been tested using sample holes within different sections of the workpiece.” (Highlighted in red color)

33-34

Abbreviation table: Capital letters… e.g., CNC: Computer Numerical Control, etc.

Thanks for your comment. The amendment in the Abbreviation table had change highlighted with red color.

---

---

References: we invite authors to consider:

-  Jbira et al., Evaluation of the Algorithmic Error of New Specification Tools for an ISO 14405-1:2016 Size, DOI:10.1115/DETC2018-85669, 2018, Conference: ASME 2018 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference

-  Jianfeng, G.,Yang, J., An Iterative Procedure for Robust Circle Fitting. Communications in Statistics - Simulation and Computation, 2018, 0(0),1–8.

-  Nurunnabi, Abdul, Yukio Sadahiro, Debra F. Laefer. 2018. Robust Statistical Approaches for Circle Fitting in Laser Scanning Three-Dimensional Point Cloud Data., Pattern Recognition 81, 417–31.

-  Nurunnabi, Ab., Yukio S., Roderik L., Robust Cylinder Fitting in Three-Dimensional Point Cloud Data. Int Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2017, 42(1W1), pp.63–70.

-  Rhinithaa, PT., Selvakumar, P., Sudhakaran,N., Anirudh,V., Lawrence K.D. Comparative Study of Roundness Evaluation Algorithms for Coordinate Measurement and Form Data, Precision Engineering 2018,51: pp.458–67.

-  Sui, W. , D. Zhang, Four Methods for Roundness Evaluation. Physics Procedia, 2012. 24, pp. 2159-2164.

-  Xiuming, L., Z. Jingcai,  L. Hongqi, Determination of the Minimum Zone Circle Based on the Minimum Circumscribed Circle. Measurement Science and Technology, 2014, 25.

-  Zhao, Z., Xi JianPu., Zhao X., Zhang, G., Shang ,M., Evaluation of the Calculated Sizes Based on the Neural Network Regression, Mathematical Problems in Engineering, 2018,pp.11.

We appreciate your comment regarding the references. We have added these references to strengthen the study reflection of the roundness circle and make the article more reliable and efficient. The references .addressed as [36,37,40, 43-47] was highlighted in red color.

144-146,

151-154,

163-171

Line 650, REF #7, ‘and’ ??????

Thanks for your privileged comment. The amendment had been changed and highlighted with yellow color.

650

647

Eq. 1 and E1. 2: Rmax=R1 ????? Italic form (lines 199-201), etc.

Thanks for your great comment. The amendment had been changed and highlighted with red color.

199-201

224-225

Line 185: The MZC method uses two concentrical circles to determine

Thank you for your comment. We have added the sentences to make the statement clearer” The MZC method determines roundness error by using two reference circles. which provides a quantitative measure by comparing the actual contour with two reference circles, the Maximum Inscribed Circle (MIC) and the Minimum Circumscribed Circle (MCC).” (highlighted by red color)

185

185-211

Line 145: form errors (not error!)… form = straightness, flatness and circularity… (ISO-GPA and ASME Y14.5)

Thanks for the comment. We had an amendment to the sentence typos that were highlighted in red color.

145

156,160-161

Line 240: The Experimental Steup and Procedure, as section 3.2.1, etc.

Thanks for the fruitful comment. We had changed the small letters to capital letters of the headline. Highlighted in red color.

240

265

Table 1: line 274, CMM resolution… units?, Acc = Highly ???? Please avoid all non-quantified terms.

Thanks for the fruitful comment. We had avoided the non-quantified data from Table on and added the units  Highlighted in red color.

274

299

In section 4.4 (line 384), appendix A… ?????? No Appendix in the paper!

We appreciate your comments.  We added the appendix as a supplementary file that shows the code of the image detection and the implementation of the algorithms.

384

408

Line 531: …roundness circle values…

Thanks for your comment. We had change the sentence” the surface feature diameter of the roundness circle values of Example 1 Part 21file based ISO14649, developed using the 3SMVI vision system in twelve points.” (Highlighted in red color)

531

555-557

Table 2: Dia. 30 mm (not 22mm)

We appreciate your comment. We had amended the typos to Dai.30mm

561

Figure 11: remove lines (green) between angle in degree… What is the uncertainty pf each measure?

 We appreciate your comment. We have deleted the green lines. Therefore, the uncertainty of the measurement in our model is +/- 0.0005m.

565

Figure 13 (line 550)… No figure 12????)

Thanks for the comment. We had changed the typo from figure 13 to 12. Highlighted in red color.

550

575

Figure 13 (line 550), error bar?... +/-u or +/- 2u, C.I.?...  Bleu lines between images?!!!

Thanks for the great comment. We have deleted the blue line between the images. (Highlighted in red color)

550

588

Table 4: confusion!!!! 3SMVI <-> CCM… same values as Table 2 and Table 3… can be reduced

Thanks for your fruitful comment. We have been deleted the table 4 from the manuscript which intend to be

558

----

Table 3: where is points A, B, C, etc.?

Thanks for your comment. We have addressed points A, B, and C on the surface feature of the model of the workpiece shown in Figure 2. Highlighted in red color.

294

Line 560: Table Error!!! In addition, how we guarantee the use the same ‘c’ (center) for all measurements images?

Thanks for your comment. We had an amendment name error in Table 5 previously and we have changed the typo to the same table that becomes Table 4. (highlighted in red color) …

Furthermore, the use of computer vision techniques created specifically for hole or circle measurement is the most efficient way to guarantee the consistent use of the same "c" (centre) for measurements in all images. We can consistently reference the same 'c' point across all measurements by implementing powerful algorithms that can automatically detect and precisely locate the centres of these features in each image, ensuring accuracy and reliability in our analyses.

560

583

Figure 14… Number of point -> 3SMVI and CMM,

We appreciate your privileged comment. The number of points means the errors between systems 3SMVI and CMM and showed the deference errors of the two systems which were 7.8 micro meters.

587

Response to Reviewer 2

Questions

Comments

Previous Line No.

New line No.

1.    Author redraws the figure 6. It is not clear.

Thanks for your fruitful comment. We had to change Fig. 6 due to the less clarity which is highlighted in red color.

424

2.    Author mentioned Image analysis and Segmentation, but no figure or diagram is given for explanation of Segmentation.

Thank you for your insightful feedback. In Figure 10, we present the model image in different formats: binary, grayscale, along with a version subjected to Gaussian blur, and another demonstrating the dilation process. These images are essential components of our discussion, illustrating the algorithms employed for analysis and segmentation."

549

3.    In this work what is edges detection, which method is used. Explain which method is better.

We have used edge detection techniques in image processing such as the Sobel Operator, which calculates image intensity gradients, and the Canny Edge Detector, a multi-stage algorithm known for accuracy and noise reduction. These methods are effective for detecting horizontal and vertical edges in images.

380-382

4.    How author do the Calibration from Camera-to-Workpiece.

we appreciate your thoughtful comments. We recognize and value your point about the calibration from the camera to the workpiece. The 3SMVI interface platform, on the other hand, is made to take a number of important factors into account when determining the pixel size in both the x and y axes. One of these parameters is the expanded model's actual size, which is taken from the STEP-NC Example 1 Part 21 file. It's also important to remember that the program takes the calibration of the camera system into account. The relationship between the camera's imaging capabilities and the dimensions of the world it captures is established through calibration. The region of interest (ROI), which designates the area of the image that is relevant for analysis, is specifically discussed. Using camera calibration and understanding of the ROI, the program ensures accurate and precise pixel-to-real-world unit conversion, which is vital for various applications in image analysis and measurement

472-476

5.    What is the reason matter of Conclusion and Future Research is highlighted.

We appreciated your comment. Therefore, it becomes clear that the "Conclusion" and "Future Research" sections serve as a well-structured way to not only conclude the present research but also to emphasize its role in advancing ongoing scientific exploration given the continuous advancements in this emerging field and its significance for the industrial sector within the context of Industry 4.0.

----

----

6.    Make a comparison table of your results with the literature work.

Thanks for your comment. We have the table 4 that concluded the comparison of the  errors between two system 3SMVI and CMM.

580

7.    Explain the novelty of this work.

Thanks for your great comment. We have added the novelty of this work in the conclusion part the novelty lies in seamlessly integrating IoT, computer vision, and image processing, enhancing industrial roundness assessment. IoT facilitates real-time data exchange for holistic object roundness analysis. A developed 3SMVI system, within IoT and CMM, achieves high accuracy. Emphasizing uncertainty assessment and advanced algorithm integration distinguishes this research.

598-609

Response to Reviewer 3

Questions

Comments

Previous Line No.

New line No.

The main concerns are given as follows:

There are grammatical and punctuation errors in the paper. The authors require a native speaker to proofread. The authors can use the professional version of the Grammarly system.

Thanks for your privileged comment. We had sent the manuscript for proofread.

-----

----

The authors should review more studies by 2023.

Thank you for your fruitful comment. We added the new articles recently published in 2023 to strengthen the study reflection of the roundness circle and make the article more reliable and efficient shown in Ref. [34, 36, and 49]. (highlighted in yellow color)

----

738,742,

777

The research methodology should be designed in Section 2.

We appreciate your great comment. We previously mentioned the design of the methodology in the manuscript end part of the introduction and we modified some sections.(Highlighted in red color)

131-132

 The title number of section 2 and other sections (Materials and Methodology) should be changed to 3.

Thanks for your comment. The amendment of the errors and typos of numbering had changed to follow the structure of the manuscript. (highlighted with red color).

---

251

The comparison between the current study and other studies should be considered in the Result and Discussion section.

We appreciate your comment regarding the references. We have deleted Table 4 due to the repeated table and extended Tables 2 and 3 for errors of round circles in each system as a vision system and coordinate measuring machine. Furthermore, the comparison of the error data of the two systems is intended in the discussion section shown in Table 4.

558,573,  583

The conclusion section should be explained more clearly and simply

Thanks for your valuable comment. The amendment had been changed in the conclusion to make it clear for the reader. (highlighted with yellow color)

-----

592-610

Response to Reviewer 1

Questions

Comments

Previous Line No.

New line No.

"I recommend the article for publication

after removing typos such as in the description of the x-axis in Fig. 13, as

well as e.g. improving the appearance of equation (1)."

Thanks for your fruitful comment. The amendment has been changed to the fig.13. (highlighted in red color)

Thanks for your comment. We had changed the equation 1 with clear appearance. (highlighted in red color)

562

211

562

211