All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
The authors have improved the work accordingly. This version addressed the concerns of the reviewers. It can be accepted.
[# PeerJ Staff Note - this decision was reviewed and approved by Jyotismita Chaki, a PeerJ Section Editor covering this Section #]
Previous correction notes have been corrected and are Clear and unambiguous, professional English is used throughout.
Previous note has been corrected, Original primary research in Aims and Scope of the journal.
Previous note has been corrected, Impact and novelty not assessed. Meaningful replication encouraged where rationale & benefit to literature is clearly stated.
Please add the experiments requested by Reviewer 2 and check the manuscript accordingly.
The authors have responded to my comments and revised the manuscript. The author should further check the manuscript to ensure that the descriptions are compliant and the formatting is correct.
The authors have provided some responses. From a computer science perspective, I still recommend that the authors add experiments on the selection of some important parameters.
Thanks to the authors for providing the specific data. The essential information in this part should also be incorporated into the manuscript.
The paper is clearly written and includes all the references to cover the background comprehensively
The paper fits the journal's scope. The problem is clearly framed, and a rigorous investigation is carried out. The methods are described in detail.
Conclusions are supported by results and the validity of findings is discussed.
The authors addressed all the raised issues
The authors have addressed some issues proposed by the reviewers. However, there are some further comments and suggestions. Please continue to improve the article accordingly.
1. It seems that the title Results is missing before the "A. Experiments in the identification section".
2. The statement that deep learning "requires large-scale labelled datasets (e.g, 100,000 images)" is too exaggerated. Thousands of examples can be enough, as witnessed by the references included in the manuscript.
3. The contribution requires clarification. The authors state that:
"The novelty of this system lies in applying the SIFT-Canny-AdaLAM algorithm to microscopic ceramic images, addressing the limitations of traditional ceramic recognition methods, such as slow processing speed and poor matching accuracy"
SIFT, Canny, and AdaLAM are well-known and established computer vision algorithms. It's a novel combination? Or the novelty is the specific application of this algorithm combination to the domain of "microscopic ceramic images"?
In both cases, the novelty should be better articulated in terms of how their specific combination and application solve the domain's challenges.
4. In row 153, the authors say that "extracted feature vector set is transferred to the blockchain storage module" then in 156: "We first store the high-cost, high-volume data on IPFS, which generates a hash. This hash, along with the low-volume data, is then stored on the blockchain". This can create confusion. I suggest to remove the word blockchain in the first sentences since this inconsistency in the order of information could lead to confusion.
5. Regarding the experimental evaluation, while the dataset is of 10,000 micrographs from 20 distinct ceramic artifacts, the specific experiments detailed (e.g., '10 pairs of local micrographs from the same folder' for initial performance comparison, and '100 sets of microscopic images from different ceramics' for false verification) utilize a relatively small subset of it. How were these specific 10 pairs (or the 100 sets) selected? Were they chosen randomly, or were there specific criteria to ensure they were representative of the overall dataset's variability and challenges? Given that the algorithms are not learning-based and thus don't require a separate training/testing split, it would also be insightful to understand why a more extensive evaluation on a larger portion, or even the entirety, of the dataset was not conducted to fully leverage the available data and further strengthen the statistical confidence in the reported performance metrics.
Please address the reviewer comments, especially the major concerns raised by reviewer 1.
**PeerJ Staff Note:** Please ensure that all review and editorial comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
**Language Note:** The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Alternatively, you should make your own arrangements to improve the language quality and provide details in your response letter. – PeerJ Staff
- Related work is too broad; focus more on ceramics and blockchain integration.
- No clear comparison of existing methods vs. proposed methods—add a summary table.
- The explanation of IPFS and blockchain feels too descriptive, less analytical.
- Grammatical issues throughout—proofreading is needed for clarity and readability.
- Algorithm steps lack clarity—pseudocode should be more concise.
- The rationale for combining SIFT-Canny-AdaLAM is weak—justify better.
- No discussion on computational cost—add performance benchmarks.
- The smart contract section lacks technical depth—expand the implementation details.
- The description of parameter tuning (Canny thresholds) is too shallow.
- Figures lack proper captions and need clearer labeling.
- MSE and precision are reported but not analyzed critically—add interpretations.
- No statistical validation (e.g., confidence intervals) for the results.
- The comparison with existing methods focuses only on accuracy, including speed and robustness.
Abstract
- Abstract lacks clear structure—focus on methods, findings, and results.
- Numerical results should be highlighted to show effectiveness clearly.
Introduction
- Some background info is irrelevant (e.g., IoT history). Shorten it.
- The research gap is vague—clarify what previous studies missed.
- Citations lack consistency; some old references could be updated.
Literature Review
- Related work is too broad; focus more on ceramics and blockchain integration.
- No clear comparison of existing methods vs. proposed methods—add a summary table.
- The explanation of IPFS and blockchain feels too descriptive, less analytical.
Discussion
- The impact of using blockchain is not quantified—discuss storage and cost overhead.
- The limitations section is too short—add challenges in real-world implementation.
1. The manuscript generally uses professional English, but a few sections could benefit from clearer phrasing to enhance reader comprehension. The language should be improved for revision.
2. The introduction provides a solid background, but more references (e.g., works in 2024) to relevant recent literature should be included to contextualize the findings better.
3. The research question is clearly defined and engages with a significant issue in ceramics identification. The submission introduces SIFT and blockchain technologies, and the related technologies are described adequately, but should include more details on their implementation to ensure reproducibility.
4. Including more statistical analyses or validation studies to reinforce claims made regarding the method's effectiveness would be beneficial.
5. The quality of the figures could be improved; it is better to use vector figures to read and understand.
6. It is better to consider adding statistical tests or methodologies that validate the robustness of the results.
1. The manuscript falls within the aims and scope of the journal, focusing on innovative methodologies for ceramic identification and traceability, which are relevant to both materials science and computer applications in art transactions.
2. The research question is well-defined and meaningful. The submission focuses on ceramic art transactions by exploring the integration of SIFT (Scale-Invariant Feature Transform) with blockchain technology for improved identification and traceability.
3. The authors assert that their proposed method could significantly improve these areas.
4. This submission emphasizes the development of a new algorithm (SIFT-Canny-AdaLAM) designed to enhance feature extraction and matching processes.
1. While the findings are presented compellingly, the authors should explicitly discuss how their results contribute to the field and compare them against existing methods quantitatively.
2. The impact and novelty of the findings are not thoroughly assessed within the manuscript.
3. All underlying data appear to be provided, as per the journal's standards. However, the authors should ensure these data are made available in a recognized discipline-specific repository, along with a description of how the data were collected and processed.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.