Nondestructive Evaluation of Heritage Object Coatings with Four Hyperspectral Imaging Systems

Round 1

Reviewer 1 Report

The article convincingly presents the advantages of using multispectral images for the analysis of paintings .
In fact, unlike the spectroscopic techniques currently used, this technique is totally non-invasive and, above all, allows as to have an overall view of an entire surface, thus favoring continuous and low-cost monitoring of the artworks.
The results of PCA analysis are interesting and well exposed.
The article also clarifies very well the importance of integrating hyperspectral and spectroscopic analysis techniques and it presents a convincing example of acquisition of regions of interest with corresponding spectra deriving from HI (fig. 6).
Therefore it deserves to be published, just some comments:
1. It does not seem to me appropriate to start a paper on cultural heritage with a purely economic indication of value ( 40-41)
2. In the analysis of the PCs and in the related figures (2-5), specify what the percentages expressed refer to (percentages of noise?)
3. In figures 2-5 clarify which value is expressed on the ordinate axis and therefore what exactly is meant by negative and positive values.
4. In the conclusions summarize the results obtained on the case study: pigments used, types of binders, resins, conservation materials, state of conservation of the support. Have these results allowed a historical-artistic evaluation of the painting? Have they directed the conservative choices?
It seems very interesting to me that, by working with different wavelength ranges, the painting can be analyzed stratigraphically: from the surface pigments to the physical characteristics of the support. This possibility of stratigraphic analysis could be highlighted and deepened in the future.

Author Response

Reviewer 1

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

English language and style

( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style

Comments and Suggestions for Authors

Dear Reviewer 1

Thank you very much for your time in reviewing our manuscript as well as for your valuable comments. We have deeply discussed your suggestions and modified the manuscript accordingly. Please find below our responses to all points raised in your Review. We do sincerely hope that we correctly interpreted your concerns and our modifications (answers) fulfill your requirements.

The article convincingly presents the advantages of using multispectral images for the analysis of paintings .
In fact, unlike the spectroscopic techniques currently used, this technique is totally non-invasive and, above all, allows as to have an overall view of an entire surface, thus favoring continuous and low-cost monitoring of the artworks.
The results of PCA analysis are interesting and well exposed.
The article also clarifies very well the importance of integrating hyperspectral and spectroscopic analysis techniques and it presents a convincing example of acquisition of regions of interest with corresponding spectra deriving from HI (fig. 6).
Therefore it deserves to be published, just some comments:

Thank you for your kind words. We did our very best to prepare our report as clear and comprehensive as possible.

1. It does not seem to me appropriate to start a paper on cultural heritage with a purely economic indication of value ( 40-41)

We agree with your, as well as other Reviewers' comment. The use of strictly economic figures in regard to cultural heritage research was inappropriate, especially at the opening of the introduction chapter. The sentence was removed in the revised manuscript.

1. In the analysis of the PCs and in the related figures (2-5), specify what the percentages expressed refer to (percentages of noise?)

The percentage values correspond to the variance explained by each principal component. We clarified this in the note in the figure caption.

1. In figures 2-5 clarify which value is expressed on the ordinate axis and therefore what exactly is meant by negative and positive values.

Graphs in figures 2-5 correspond to the loadings of the Principle Component Analysis (PCA) model. These are unit-less values that are used by the model for the reconstruction of the original spectrum. In that case, the number of selected loadings are multiplied by corresponding scores and summed. The difference between reconstructed by the model and the original spectrum is defined as residual. The overall objective for PCA modeling is to minimize residuals and to assure that only noise data are included there. We included the first three components for the PCA model in our report that explained from 46.3% (MWIR) to 89.5% (SWIR) of variance. The sign of the loading does not have any physical meaning as it is connected to the sign of the corresponding score. Therefore, the reconstruction is successful if a correct sign of both loading and scores are derived by the model. However, the distance from the horizontal axis has a great influence on the model and all high (both positive and negative) values of loading indicate the importance of a given variable for the reconstruction of that particular component.

1. In the conclusions summarize the results obtained on the case study: pigments used, types of binders, resins, conservation materials, state of conservation of the support. Have these results allowed a historical-artistic evaluation of the painting? Have they directed the conservative choices?

It is a very relevant comment that determines the useability of hyperspectral imaging for cultural heritage studies. The part of the research team involved in this study includes experts from the national cultural heritage institute who assisted interpretation of the results obtained. The evident limitation of the work reported here is the lack of reference data indispensable for the development of more advanced chemometric models. Even if we performed several additional analyses (not reported here), the overall performance of discriminant analysis was not yet satisfactory. We are convinced that along with further research our portfolio of reference spectra will dramatically increase as well as the capability of hyperspectral imaging systems to address all the info listed in the Reviewer’s comment. In addition, we included the following text in the Conclusion chapter that includes a summary of our findings as well as additional comments.

Line 395: “Traces of lead white, emerald green, prussian blue, and carbon-based black were identified as pigments used by the artist. Moreover, lipid and triterpenoid resins as binders were detected in the spectra. Detailed analysis of results allowed the identification of damaged areas of studied beehive panel painting, as well as hidden underdrawings in the inner layer.”

It seems very interesting to me that, by working with different wavelength ranges, the painting can be analyzed stratigraphically: from the surface pigments to the physical characteristics of the support. This possibility of stratigraphic analysis could be highlighted and deepened in the future.

We do agree with this remark and hope to further extend our research in the future focusing on the capacity for stratigraphic analysis. In addition, we included some amendments in the revised manuscript, such as:

Line 119: “Analysis of CH objects in short wave infrared (SWIR) and mid-wave infrared (MWIR) regions provide the possibility of visualizing hidden details in the inner layers of paintings, highlighting the underdrawings [37]. The stratigraphic analysis is a highly interesting approach for investigating internal layers of complex surfaces. The majority of applicable techniques, including microscopy, radiography, fluorescence, diffractometry, or spectroscopy, are considered as non-destructive [13, 38-44] and consequently suitable for the characterization of cultural heritage objects. However, the effective penetration depth of corresponding radiation depends on its nature, energy and wavelength. The operational depth of X-ray diffractometers (XRD) varies from a few to approximately 100 µm, depending on the optical density of the material and its X-ray attenuation. Scanning electron microscopy combined with Energy-dispersive X-ray spectroscopy (SEM/EDX) provides information of approximately 2 µm thick subsurface. The excitation wavelength of light in Raman spectroscopy determines the light penetrations depth. It corresponds to 0.7 µm with 532 nm source and increases to 12 µm when 785 nm laser source is irradiating the surface. Attenuated total reflectance (ATR) Fourier transform infrared spectroscopy (FTIR) is frequently used for the assessment of the physical-chemical properties of materials. The use of specific crystal, such as germanium (Ge), zinc selenide (ZnSe), thallium-bromoiodide (KRS-5), silicone (Si), or diamond (C) affects the light refractive index and influence the penetration depth of infrared toward object’s surface. In that case, it varies between approximately 0.5 and 2 µm. Much higher penetration is noticed for near infrared radiation, which can vary from 0.5 to 5 mm, again depending on the optical density of tested material and wavelength of irradiating light [45].]”

And

Line 334: “The detailed analysis of hyperspectral images, including both raw spectra inspections as well as higher degree principle components visualization, lead to the discovery of several features that were not apparent by simple panel observations, such as underdrawings, retouching, and/or pentimenti.”

Summarizing, we would like to thank you once again for all the kind comments and suggestions. We do believe that we addressed all of these and the revised manuscript may fulfill your expectations.

With sincerely regards,

Jakub Sandak, on behalf of authors

Reviewer 2 Report

The Authors present a quite extensive work for the examination of very peculiar, local and interesting artworks that can be assumed as an example spectroscopic analysis of artworks for a better understanding and for conservative purposes.

While the paper is interesting and claims for a publication, I recommend an extensive correction of concepts and typos that are necessary to amend the text. In the following I’ll try to give a tentative list of needed corrections and comments on specific points.

1. Readability and correctness can be increased by remembering to define every acronym before using it, in particular in the Abstract (e.g., insert “(HI)” at the end of definition, after “imaging” in line 24). Also, “CH”, while not previously defined, is used in line 51, while it is defined in line 55. Please, check similar problems overall along the text.
2. I’m in complete disagreement with the aim and the numbers given just in the very first sentence (line 40-41). After a typo (“billon” instead of “billion”), there is the mention of three million people involved in art worldwide. But, as the quite commercial reference claims, those are numbers related only to the art market, therefore dealing with the very small part of objects that can be traded, sold, and freely acquired. It means that the cited reference is related to a very small part of the whole “Cultural heritage” patrimony that nations worldwide and, for sure, in Europe, have to keep safe to save memories of their past, including paintings, sculptures, frescos and ruins that cannot be sold, or moved, and can be visible only in specific areas or museums. The economic impacts to be mentioned would be instead related to the complete cultural assets at national levels, including tourism activities, therefore having much more important economical and occupational impacts. I think, at the end, that any scientific instrument able to help CH conservation can have much more importance than the number cited in Ref. 1 (that reference in line 346 retains a not-so-interesting information “accessed on ….”). I suggest the Authors to modify this very first part including more numbers, more references, or a different one, and reformulate the concept in a more general way.
3. Lines 58-60. I agree, HI provides a map of spectroscopic properties, but shares with VNIR spectroscopy the ability to be a non-contact technique, while the sentence there assigns a double advantage of HI over VNIR spectroscopy, that it is not true at all. Please amend it.
4. You could move the reference [10] from line 62 to line 61, just after “Roselli & Testa” mention. There is a similar situation at line 81 for reference [13], that has to be moved at the beginning of line 80, just after “et al.”, where a full stop is also missing, I think.
5. Lines 107,108 and 109. There is a misplaced full stop before “Although”, I suppose.
6. Line 131. Table. “This is a table ….” Check it.
7. Lines around 170, and everywhere. I think that there is a general, although common, misconception of “color” vs. wavelength. A hyperspectral system does not see “colors”, because simply realizes a pile of monochromatic images, each one at a very precise wavelength. A color is often a complicated superposition of pure wavelengths, while it can be “projected” over an RGB coordinate systems, simulating the behavior of our eye (note that the word “color” has a very precise meaning only for the very reduced spectral interval available to our eyes). But such bad misconceptions are very common, unfortunately. Consider this comment, if you can.
8. At the end, I would like to add a final comment. To fairly measure the ability of the Authors in discovering the precise chemical components of the interesting artworks under consideration, for which different compositions and aging factors are intermixed in an unknown way, some measurements of specially prepared standard samples could be necessary (actually, as declared by the Authors as the future lines of work). Otherwise, the performed analysis is simply qualitative, being able to identify parts of the surface with similar characteristics. Note, however, that, as also correctly mentioned by the Authors, surface details (e.g., particle dimension) can alter the optical response in such a way that similar parts (on compositional wise) appear different, or, on the contrary, different zones of different chemical composition can be seen as similar. All that can be true, in particular, in VNIR spectral part, due to the strong scattering dependence on the particle sizes, or on the strong wavelength dependence of the index of refraction of the involved painting layers, much less for MWIR regions. Also, a performances evaluation against more chemical oriented tools (e.g., Raman spectroscopy) could be clearly stated. Therefore, a more detailed critical conclusion can be added to clearly state what it is obtained, what can be obtained, and the pro and cons of the used techniques. For instances, some comments and figures about the morphologic depth accessible to spectroscopy (also vs different wavelengths) in comparison with other techniques (e.g., Raman, XRay, …) is completely missing, including a cross-sectional information of those paintings, while maybe already available in the current literature.

Author Response

Reviewer 2

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

English language and style

( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style

Comments and Suggestions for Authors

Dear Reviewer 2

Thank you very much for your time in reviewing our manuscript as well as for your valuable comments. We have deeply discussed your suggestions and modified the manuscript accordingly. Please find below our responses to all points raised in your Review. We do sincerely hope that we correctly interpreted your concerns and our modifications (answers) fulfill your requirements.

The Authors present a quite extensive work for the examination of very peculiar, local and interesting artworks that can be assumed as an example spectroscopic analysis of artworks for a better understanding and for conservative purposes.

While the paper is interesting and claims for a publication, I recommend an extensive correction of concepts and typos that are necessary to amend the text. In the following I’ll try to give a tentative list of needed corrections and comments on specific points.

Thank you for your kind general assessment and the following suggestions. We did our best to address each and to revise the text as in the re-submitted manuscript.

1. Readability and correctness can be increased by remembering to define every acronym before using it, in particular in the Abstract (e.g., insert “(HI)” at the end of definition, after “imaging” in line 24). Also, “CH”, while not previously defined, is used in line 51, while it is defined in line 55. Please, check similar problems overall along the text.

The definition of abbreviations was carefully inspected and provided for all missing cases.

1. I’m in complete disagreement with the aim and the numbers given just in the very first sentence (line 40-41). After a typo (“billon” instead of “billion”), there is the mention of three million people involved in art worldwide. But, as the quite commercial reference claims, those are numbers related only to the art market, therefore dealing with the very small part of objects that can be traded, sold, and freely acquired. It means that the cited reference is related to a very small part of the whole “Cultural heritage” patrimony that nations worldwide and, for sure, in Europe, have to keep safe to save memories of their past, including paintings, sculptures, frescos and ruins that cannot be sold, or moved, and can be visible only in specific areas or museums. The economic impacts to be mentioned would be instead related to the complete cultural assets at national levels, including tourism activities, therefore having much more important economical and occupational impacts. I think, at the end, that any scientific instrument able to help CH conservation can have much more importance than the number cited in Ref. 1 (that reference in line 346 retains a not-so-interesting information “accessed on ….”). I suggest the Authors to modify this very first part including more numbers, more references, or a different one, and reformulate the concept in a more general way.

We do agree with your, as well as other Reviewers' comment. The use of strictly economic figures in regard to cultural heritage research was inappropriate, especially at the opening of the introduction chapter. The sentence was removed in the resubmitted manuscript and the overall chapter revised. There are several additional references included in the revised manuscript that we do hope will better provide the context of our research, without highlighting economic aspects of the research in cultural heritage.

Line 40: “Imaging techniques used in the conservation of cultural heritage objects play an important role in authentication, documentation, assessment of materials composition, detection of past conservation, evaluation of object state and programming further restoration actions. All of these activities relayed in the past entirely on the personal experience of highly trained experts [1]. However, with the development of alternative diagnostic techniques modern “high-tech” instruments become more frequently used in the field of art analysis. The progress in the technology of high-resolution silicon charge-coupled device (CCD) has greatly improved the availability and performance of digital imaging systems [2] in diverse application fields [3,4]. The revolution of optoelectronics provided new solutions for in-situ measurements, broad application of miniaturized portable devices [5-7] or smartphones [8-9], offering optimal performance with relatively low costs.”

1. Lines 58-60. I agree, HI provides a map of spectroscopic properties, but shares with VNIR spectroscopy the ability to be a non-contact technique, while the sentence there assigns a double advantage of HI over VNIR spectroscopy, that it is not true at all. Please amend it.

The sentence was corrected as suggested:

Line 71: “The biggest advantage of HI over Vis-NIR spectroscopy is that imaging allows detailed analysis of the spatial distribution of particular pigments. Though, both techniques allow measurements in a non-invasive manner, excluding any possible physical contact with the surface of the CH objects.”

1. You could move the reference [10] from line 62 to line 61, just after “Roselli & Testa” mention. There is a similar situation at line 81 for reference [13], that has to be moved at the beginning of line 80, just after “et al.”, where a full stop is also missing, I think.

Both references are corrected as suggested.

1. Lines 107,108 and 109. There is a misplaced full stop before “Although”, I suppose.

The sentence is corrected, as suggested.

1. Line 131. Table. “This is a table ….” Check it.

Thank you very much, this was clearly omitted when creating the final version of the manuscript by using the template. The proper caption for table 1 is provided:

Line 178: “Table 1. Spectroscopic characteristics of hyperspectral cameras used for investigation”

1. Lines around 170, and everywhere. I think that there is a general, although common, misconception of “color” vs. wavelength. A hyperspectral system does not see “colors”, because simply realizes a pile of monochromatic images, each one at a very precise wavelength. A color is often a complicated superposition of pure wavelengths, while it can be “projected” over an RGB coordinate systems, simulating the behavior of our eye (note that the word “color” has a very precise meaning only for the very reduced spectral interval available to our eyes). But such bad misconceptions are very common, unfortunately. Consider this comment, if you can.

We do agree with this very important remark. Perhaps it is due to our limited proficiency in composing fully correct English sentences as well as limited vocabulary. The manuscript was peer-reviewed by the native English Editor, but still, some imperfections seem to be present. We reviewed the text before re-submission with a particular focus on the proper use of words “colour” and “wavelength”. Hope it is now improved.

1. At the end, I would like to add a final comment. To fairly measure the ability of the Authors in discovering the precise chemical components of the interesting artworks under consideration, for which different compositions and aging factors are intermixed in an unknown way, some measurements of specially prepared standard samples could be necessary (actually, as declared by the Authors as the future lines of work). Otherwise, the performed analysis is simply qualitative, being able to identify parts of the surface with similar characteristics. Note, however, that, as also correctly mentioned by the Authors, surface details (e.g., particle dimension) can alter the optical response in such a way that similar parts (on compositional wise) appear different, or, on the contrary, different zones of different chemical composition can be seen as similar. All that can be true, in particular, in VNIR spectral part, due to the strong scattering dependence on the particle sizes, or on the strong wavelength dependence of the index of refraction of the involved painting layers, much less for MWIR regions. Also, a performances evaluation against more chemical oriented tools (e.g., Raman spectroscopy) could be clearly stated. Therefore, a more detailed critical conclusion can be added to clearly state what it is obtained, what can be obtained, and the pro and cons of the used techniques. For instances, some comments and figures about the morphologic depth accessible to spectroscopy (also vs different wavelengths) in comparison with other techniques (e.g., Raman, XRay, …) is completely missing, including a cross-sectional information of those paintings, while maybe already available in the current literature.

The above critics is in majority of the context accurate and we do agree with such assessment. The research performed till now has important limitations, as properly identified by the Reviewer. However, we intentionally limited “overinterpretations” of our findings to assure as high as possible credibility of our report. We do agree that there is a lot of research ongoing in the direction of hyperspectral imaging in the context of cultural heritage. We are also aware of several theoretical and experimental advances regarding stratigraphic use of diverse analytical techniques, as briefly mentioned in our manuscript. However, by considering a relatively low level of reference information available for these particular objects of our studies, we decided to limit the “extrapolation” of our “solid” scientific conclusions and to focus on the demonstration of various HI techniques. It is also important to state, that the general objective of our work was to investigate if the integration of diverse hyperspectral imaging cameras can be considered (to some extent) as a multi-sensor approach, exactly as pointed by the Reviewer. Again, there are still several methodological short breaks in the report presented, such a different resolution of each camera, difference in the variance explained by PCA models, among others. That is why the follow-up of this research is to provide an algorithm for multisensory data fusion focused on the hyperspectral imaging hypercubes as well as integration with an alternative to hyperspectral imaging techniques. Again, as the research team involved in this project is relatively broad, we are doing our best to combine diverse expertise and provide some innovation in this field.

Finally, following the suggestion of the Reviewer we included the following paragraph regarding penetration depts of diverse radiation waves:

Line 121: “The stratigraphic analysis is a highly interesting approach for investigating internal layers of complex surfaces. The majority of applicable techniques, including microscopy, radiography, fluorescence, diffractometry, or spectroscopy, are considered as non-destructive [13, 38-44] and consequently suitable for the characterization of cultural heritage objects. However, the effective penetration depth of corresponding radiation depends on its nature, energy and wavelength. The operational depth of X-ray diffractometers (XRD) varies from a few to approximately 100 µm, depending on the optical density of the material and its X-ray attenuation. Scanning electron microscopy combined with Energy-dispersive X-ray spectroscopy (SEM/EDX) provides information of approximately 2 µm thick subsurface. The excitation wavelength of light in Raman spectroscopy determines the light penetrations depth. It corresponds to 0.7 µm with 532 nm source and increases to 12 µm when 785 nm laser source is irradiating the surface. Attenuated total reflectance (ATR) Fourier transform infrared spectroscopy (FTIR) is frequently used for the assessment of the physical-chemical properties of materials. The use of specific crystal, such as germanium (Ge), zinc selenide (ZnSe), thallium-bromoiodide (KRS-5), silicone (Si), or diamond (C) affects the light refractive index and influence the penetration depth of infrared toward object’s surface. In that case, it varies between approximately 0.5 and 2 µm. Much higher penetration is noticed for near infrared radiation, which can vary from 0.5 to 5 mm, again depending on the optical density of tested material and wavelength of irradiating light [45].”

Summarizing, we would like to thank you once again for all the kind comments and suggestions. We did our best to address all of these and do hope that the revised manuscript may fulfill your expectations.

With sincerely regards,

Jakub Sandak, on behalf of authors

Reviewer 3 Report

The manuscript reports on the use of "hyperspectral" cameras operating in different spectral ranges (400-1000, 900-1700, 1000-2500 and 1550-5950) for the study of cultural heritage objects. This area is growing very quickly and the review is interesting. However, the science is not new since mobile instruments in particular NIR have been used for decades in the field of the army, space, agriculture / agronomy and medicine thanks to the great development of the miniaturization of CCD detectors with the production of smartphones, which makes it possible to make cameras available at moderate cost with good performance. In each wavelength range, spectrometers allow imaging.

For example, over the past 10 years, numerous articles and review articles have been published (eg Bellon et al., Appl. Spectrosc. 1993; Zhang et al., Precision agriculture, 2012; etc.). The use of mobile instruments in the cultural field is also old (Colomban et al., J. Raman Spectrosc. 2004) and the development of high resolution hyperspectral cameras documented for a decade. The articles by Delaney et al. are cited but there are many others (e.g. Kawakami et al., IEEE Conference, 2011; Groves et al., SPIE Proc., 2014; Sun et al., Scientific reports, 2015; etc.). Modifications to the software of commercially available digital cameras have already been reported (Strivay et al., Appl. Phys. A-Mater Sci. & Process. 2016; Van Hoey et al., Radiat. Prot. Dosymetry, 2015; etc.).

Authors should better present the historical context and cite a few seminal articles (cited above and others). This will give more interest and impact to the submitted manuscript.

The authors must also better explain the advantages and disadvantages of the different spectral ranges analyzed with hyperspectral cameras. The contribution to harmonic and combining band spectra differs with the nature of the phases, organic or inorganic, and not only with their quantity, etc. and the mixing of the different contributions requires the use of complex data processing to access less partial information. This point has been widely studied in agricultural and food analysis. A paragraph on data processing dealing specifically should be present (supervised or unsupervised data processing, as for example in references 36 and 37, see also Capobianco et al., Microchemical J, 2017). This part should be better emphasized.

Author Response

Reviewer 3

Open Review

(x) I would not like to sign my review report
( ) I would like to sign my review report

English language and style

( ) Extensive editing of English language and style required
(x) Moderate English changes required
( ) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style

Comments and Suggestions for Authors

Dear Reviewer 3

Thank you very much for your time in reviewing our manuscript as well as for your valuable comments. We have deeply discussed your suggestions and modified the manuscript accordingly. Please find below our responses to all points raised in your Review. We do sincerely hope that we correctly interpreted your concerns and our modifications (answers) fulfill your requirements.

The manuscript reports on the use of "hyperspectral" cameras operating in different spectral ranges (400-1000, 900-1700, 1000-2500 and 1550-5950) for the study of cultural heritage objects. This area is growing very quickly and the review is interesting. However, the science is not new since mobile instruments in particular NIR have been used for decades in the field of the army, space, agriculture / agronomy and medicine thanks to the great development of the miniaturization of CCD detectors with the production of smartphones, which makes it possible to make cameras available at moderate cost with good performance. In each wavelength range, spectrometers allow imaging.

For example, over the past 10 years, numerous articles and review articles have been published (eg Bellon et al., Appl. Spectrosc. 1993; Zhang et al., Precision agriculture, 2012; etc.). The use of mobile instruments in the cultural field is also old (Colomban et al., J. Raman Spectrosc. 2004) and the development of high resolution hyperspectral cameras documented for a decade. The articles by Delaney et al. are cited but there are many others (e.g. Kawakami et al., IEEE Conference, 2011; Groves et al., SPIE Proc., 2014; Sun et al., Scientific reports, 2015; etc.). Modifications to the software of commercially available digital cameras have already been reported (Strivay et al., Appl. Phys. A-Mater Sci. & Process. 2016; Van Hoey et al., Radiat. Prot. Dosymetry, 2015; etc.).

Authors should better present the historical context and cite a few seminal articles (cited above and others). This will give more interest and impact to the submitted manuscript.

Preparation of the introductory chapter for the majority of the scientific publications is not an easy task. The high number of very interesting papers appearing in several scientific journals is overwhelming, not to mention revolutions in the scientific instruments and data analysis solutions. Therefore, we do apologize if our original selection of reference papers was identified as insufficient. However, following the Reviewer’s suggestion, we extensively revised the introduction chapter and included 30 additional references. We have also extended the scope of introduction to include aspects pointed in the review, such as portable instruments, the evolution of spectroscopic sensors, and diverse application fields. Please refer to all changes in the revised manuscript for more details. All of these are clearly highlighted as track changes. The list of additional references includes:

1. Gavrilov, D.; Maeva, E.; Grube, O.; Vodyanoy, I.; Maev, R. Experimental Comparative Study of the Applicability of Infrared Techniques for Non-destructive Evaluation of Paintings. J. Am. Inst. Conserv. 2013, 52, 48–60
2. Casini, A.; Lotti, F.; Picollo, M.; Stefani, L.; Buzzegoli E. Image Spectroscopy Mapping Technique for Non-Invasive Analysis of Paintings. Studies in Conservation 1999, 44(1), 39-48.
3. Bellon-Maurel,V.; Vallat, C.; Goffinet, D. Quantitative Analysis of Individual Sugars during Starch Hydrolysis by FT-IR/ATR Spectrometry. Part I: Multivariate Calibration Study - Repeatibility and Reproducibility. Appl. Spectrosc. 1995, 49, 556-562.
4. Zhang, C.; Kovacs, J.M. The application of small unmanned aerial systems for precision agriculture: a review. Precision Agric. 2012, 13, 693–712 DOI 10.1007/s11119-012-9274-5.
5. Colomban, P.; Milande, V.; Le Bihan, L. On‐site Raman analysis of Iznik pottery glazes and pigments. 2004, 35(7), 527-535.
6. Catelli, E.; Sciutto, G.; Prati, S.; Valente Chavez Lozano, M.; Gatti, L.; Lugli, F.; Silvestrini, S.; Benazzi, S.; Genorini, E.; Mazzeo, R. A new miniaturised short-wave infrared (SWIR) spectrometer for on-site cultural heritage investigations Talanta 2020, 218, 121112.
7. Perri, A.; Nogueira de Faria, B.E.; Teles Ferreira, D.C.; Comelli, D.; Valentini, G.; Polli, D.; Cerullo, G.N.; Manzoni, C. A hyperspectral camera for conservation science, based on a birefringent ultrastable common path interferometer, Proc. SPIE 11058, Optics for Arts, Architecture, and Archaeology VII, 110580B, 2019 https://doi.org/10.1117/12.2525580
8. Daffara, Cl.; Marchioro, G.; Ambrosini, D. Smartphone diagnostics for cultural heritage", Proc. SPIE 11058, Optics for Arts, Architecture, and Archaeology VII, 110581K, 2019 https://doi.org/10.1117/12.2527560
9. Van Hoey, O.; Salavrakos, A.; Marques, A.; Nagao, A.; Willems, R.; Vanhavere, F.; Cauwels, V.; Nascimento, L.F. Radiation dosimetry properties of smartphone CMOS sensors. Radiat. Prot. Dosim. 2016, 168, 314–321.
10. France, F.G.; Christens-Barry, W.; Toth, M.B.; Boydston K. Advanced image analysis for the preservation of cultural heritage", Proc. SPIE 7531, Computer Vision and Image Analysis of Art, 75310E, 2010, https://doi.org/10.1117/12.839000
11. Alfeld, M.; de Viguerie, L. Recent developments in spectroscopic imaging techniques for historical paintings - A review. Spectrochim Acta B 2017, 136, 81–105
12. Groves, R.M.; Caballero, J.; Quinzan, I.; Ribes-Gómez, E. Damage and deterioration monitoring of artwork by data fusion of 3D surface and hyperspectral measurements." Proc. SPIE 9141, Optical Sensing and Detection III, 91411E, 2014, https://doi.org/10.1117/12.2051898
13. Kawakami, R.; Matsushita, Y.; Wright, J.; Ben-Ezra, M.; Tai, Y.W.; Ikeuchi, K. High-resolution hyperspectral imaging via matrix factorization. CVPR 2011, 2329-2336
14. Peery, T.R.; Messinger, D. MSI vs. HSI in cultural heritage imaging. Proc. SPIE 10768, Imaging Spectrometry XXII: Applications, Sensors, and Processing, 107680G, 2018, https://doi.org/10.1117/12.2320671
15. Raimondi, V.; Conti, C.; Lognoli, D.; Palombi L. Latest advancements in fluorescence hyperspectral lidar imaging of the cultural heritage. Proc. SPIE 9065, Fundamentals of Laser-Assisted Micro- and Nanotechnologies, 90650Y 2013, https://doi.org/10.1117/12.2052995
16. Capobianco, G.; Bracciale, M.P.; Sali, D.; Sbardella, F.; Belloni, P.; Bonifazi, G.; Serranti, S.; Santarelli, M.L.; Cestelli Guidi, M. Chemometrics approach to FT-IR hyperspectral imaging analysis of degradation products in artwork cross-section. Microchem. J. 2017, 132, 69–76.
17. Sun, M. et al. What’s Wrong with the Murals at the Mogao Grottoes: A Near-Infrared Hyperspectral Imaging Method. Sci. Rep. 5, 2015, 14371; doi: 10.1038/srep14371
18. Kogou, S.; Lee, L.; Shahtahmassebi, G.; Liang, H. A novel methodology for the automatic analysis of large collections of paintings. Proc. SPIE 11058, Optics for Arts, Architecture, and Archaeology VII, 110580Q, 2019, https://doi.org/10.1117/12.2527611
19. Bai, D.; Messinger, D.W.; Howell, D. A pigment analysis tool for hyperspectral images of cultural heritage artifacts", Proc. SPIE 10198, Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XXIII, 101981A, 2017, https://doi.org/10.1117/12.2261852
20. Simon, C.; Huxhagen, U.; Mansouri, A.; Heritage, A.; Boochs, F.; Marzani, F.S. Integration of high-resolution spatial and spectral data acquisition systems to provide complementary datasets for cultural heritage applications, Proc. SPIE 7531, Computer Vision and Image Analysis of Art, 75310L, 2010, https://doi.org/10.1117/12.838891
21. Hayem-Ghez, A.; Ravaud, E.; Boust, C.; Bastian, G.; Menu, M.; Brodie-Linder, N. Characterizing pigments with hyperspectral imaging variable false-color composites. Appl. Phys. A Mater. Sci. Process. 2015, 121, 939–947
22. Strivay, D.; Clar, M.; Rakkaa, S.; Hocquet, F. Development of a translation stage for in situ noninvasive analysis and high-resolution imaging, Appl. Phys. A 2016, 122, 1–5, http://dx.doi.org/10.1007/s00339-016-0476-y
23. Padalkar, M.V.; Pleshko, N. Wavelength-dependent Penetration Depth of Near Infrared Radiation into Cartilage. Analyst 2015, 140, 2093-2100 doi: 10.1039/c4an01987c
24. Tonazzini, A.; Salerno, E.; Abdel-Salam, Z.A.; Harith, M.A.; Marras, L.; Botto, A.; Campanella, B.; Legnaioli, S.; Pagnotta, S.; Poggialini, F.; et al. Analytical and mathematical methods for revealing hidden details in ancient manuscripts and paintings: A review. J. Adv. Res. 2019, 17, 31–42.
25. Kubik, M. Hyperspectral Imaging: A New Technique for the Non-Invasive Study of Artworks. In: Physical techniques in the study of art, archaeology and cultural heritage, Creagh, D.; Bradley, D. (Eds). Elsevier, Amsterdam, The Netherlands, 2007, Vol. 2, pp. 199-259.
26. Rosi, F.; Miliani, C.; Braun, R.; Harig, R.; Sali, D.; Brunetti, B.G.; Sgamellotti, A. Noninvasive analysis of paintings by mid-infrared hyperspectral imaging. Angew. Chem. Int. Ed. 2013, 52, 5258–5261.
27. Dooley, K.A.; Conover, D.M.; Glinsman, L.D.; Delaney, J.K. Complementary standoff chemical imaging to map and identify artist materials in an early Italian Renaissance panel painting. Angew. Chem. Int. Ed. 2014, 126, 13995–13999.
28. MacLennan, D.; Trentelman, K.; Szafran, Y.; Woollett, A.T.; Delaney, J.K.; Janssens, K.; Dik, J. Rembrandt's An Old Man in Military Costume: Combining hyperspectral and MA-XRF imaging to understand how two paintings were painted on a single panel. J. Am. Inst. Conserv, 2019, 58, 54-68.
29. Delaney, J.K.; Zeibel, J.G.; Thoury, M.; Littleton, R.O.Y.; Palmer, M.; Morales, K.M.; de la Rie, E.R.; Hoenigswald, A.N.N. Visible and infrared imaging spectroscopy of Picasso’s Harlequin musician: Mapping and identification of artist materials in situ. Appl. Spectrosc. 2010, 64, 584–594.
30. Gabrieli, F.; Dooley, K.A.; Facini, M.; Delaney, J.K. Near-UV to mid-IR reflectance imaging spectroscopy of paintings on the macroscale, Scie Adv, 2019, 5, eaaw7794, 10pp.

The authors must also better explain the advantages and disadvantages of the different spectral ranges analyzed with hyperspectral cameras. The contribution to harmonic and combining band spectra differs with the nature of the phases, organic or inorganic, and not only with their quantity, etc. and the mixing of the different contributions requires the use of complex data processing to access less partial information. This point has been widely studied in agricultural and food analysis. A paragraph on data processing dealing specifically should be present (supervised or unsupervised data processing, as for example in references 36 and 37, see also Capobianco et al., Microchemical J, 2017). This part should be better emphasized.

Considering the reviewer suggestions, we included several new references dealing with data mining and included new text within the manuscript.

Line 105: “New methods for acquisition of high-resolution hyperspectral images [25-27], or new algorithms for post-processing of spectral data [22, 28-33] are currently within the interest of engineers and researchers working in this field. Similarly, multi-sensor systems allowing automatic characterization routines of art objects with high accuracy and precision are desired [34].”

and

Line 112: “However, considering the recent development in optics and electronics, some limitations regarding detector sensitivity as well as data storage and mining might be problematic. Frequently, after preliminary research with hyperspectral imaging, multispectral cameras equipped with a set of narrow-band optical filters are designed to lower the hardware costs and increase the computation speed, while at the same time still providing fully usable information [35-36].”

Moreover, in the revised manuscript we included a more focused comparison and objectively summarized of advantages and disadvantages of diverse complementary analytical techniques. It followed yours, as well as other reviewers' suggestions. Please find all the related modifications clearly marked as track changes in the re-submitted manuscript.

Summarizing, we would like to thank you once again for all the valuable comments and suggestions. We did our best to address all of these and do hope that the revised manuscript may fulfill your expectations.

With sincerely regards,

Jakub Sandak, on behalf of authors

Round 2

Reviewer 3 Report

All points have been addressed

Author Response

Dear Reviewer,
thank you very much, again, for your time and previous comments. Following a recommendation from Editor we carefully revised some vocabulary and grammar imperfections. We do believe that the re-submitted Manuscript can be re-considered for publication. 
With sincerely regards, 
Jakub Sandak, on behalf of Authors