Automated Defect Detection Using Threshold Value Classification Based on Thermographic Inspection

Round 1

Reviewer 1 Report

This paper is intended to use active infrared thermography, data filtering and automation to detect defects in objects. A steel plate, in this sample case. 

 

The use of non-destructive techniques such as active IRT for different sorts of defects is widely known, so any improvements in this topic is beneficial.

 

The paper is well structured, written, and scientifically sound, but several changes are needed:

 

Please cite any software and equipment used in this research.

 

Line 166: please define SNR and any acronym used in the paper the first time they are mentioned.

 

The Authors did not describe the experiment conditions in relation to possible environmental radiant sources affecting the measurement such as the operators (people) in the room, heating devices or building parts, etc. Was the survey conducted in a closed and isolated room?

 

The Results section should be named Results and Discussion, given its content and the lack of a separate discussion section.

 

Please highlight in one or two sentences the novelty of this research against other publications in the field.

 

The Authors described some limitations, but they should make sure all limitations are mentioned.

 

Please specifically describe the implications of this research.

 

The conclusions should be supported by the main quantitative data obtained in this research.

 

 

LANGUAGE

 

Lines 54, 56: Please revise the "defect defects" and "to automatic defect defects". Detect. Also, "automatically".

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

The paper reports on a study concerning the lock-in infrared thermography, that was applied to a stainless steel metal plate for automatic defect detection. A binarization processing was carried out, and a detectability analysis according to filtering was performed.

The paper is interesting, and the research steps are clearly explained. The conclusions are consistent with the results obtained.

In my opinion it would be advisable to carry out the following minor revisions before proceeding with publication.

 

1) Abstract and Keywords.

- Since authors focus on stainless steel, it would be appropriate to specify it also in the abstract and in the keywords.

2) Introduction.

- line 45. Acronym STS. Please make sure that all acronyms are written in full the first time they appear in the text.

- line 56. Please check this sentence, as "defect" is repeated twice.

3) Theory.

- lines 99-100. This claim should be supported by an explanation.

- subsection 2.2. Before applying Otsu Algorithm, it would be appropriate to explain why author have chosen to use it.

- line 115 and lines 118-119. Please check the sentences.

4) Experimental.

- Fig. 2 and Fig. 5. The characters and writings are very small, and the reading of the figures is difficult.

5) Results.

- lines 231-235. Please explain what is meant by "bwareaopen" morphology function, "strel" function, "nonholes" option. Are they functions / options of the algorithm?

Finally, please check the entire paper and correct any typos and inaccuracies.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

I have the following comments and suggestions for the authors of the article:
- Were the sample properties determined experimentally? If not, from what source (e.g. publication) the data came from.
- The authors of the article limited their work to detecting defects in the shape of circles. It is not in the shape of typical real defects. the proposed method may work for circle-shaped defects, but not useful for real defects. I am asking authors for comment.
- The main problem presented in the article are disturbances caused by uneven heating of the sample by the lamps. Such distortions are well reduced by the PCA method (especially the 1st component). Maybe this method should be used in further work.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The Authors addressed all my comments.

Author Response

Thank you so much for responding to the review.