Image-Type Data Security via Dynamic Cipher Composition from Method Libraries

Round 1

Reviewer 1 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

1. It is not clear what metrics are NPCR (%), UACI (%)
2. It is a little unclear what the dynamic multi-method encryption strategy in combination with spatial segmentatio,n is
3. How do you use spatial segmentation?
4. What new technology do you propose? What is the technological feasibility of your method? Why was this journal chosen?
5. Where are the areas of application of your method?
6. In Table 6, there is bad text in the last line.

Author Response

Comments 1:[It is not clear what metrics are NPCR (%), UACI (%)]

Response 1: [Thank you for this valuable comment. We agree that the description of the NPCR (%) and UACI (%) metrics was not sufficiently clear in the original manuscript. We have therefore revised the text to include a more detailed explanation of their meaning, importance in image encryption evaluation, and the mathematical formulas used for their computation.

]

Comments 2:[It is a little unclear what the dynamic multi-method encryption strategy in combination with spatial segmentation is]

Response 2: [We appreciate this observation. To improve clarity, we have expanded the explanation of our proposed encryption strategy in the revised manuscript (see section 2).

]

Comments 3:[How do you use spatial segmentation?]

Response 3: [Thank you for this important question. We use spatial segmentation by divding the image into 8 radial segmentation. Each segment is then encrypted independently using a method from a set that is dynamically selected. This leads to the breaking of spatial correlation , breaking the uniform encryption implementation pattern and increasing of the entropy which adds protection against attacks that are statistical or structural based.]

Comments 4:[What new technology do you propose? What is the technological feasibility of your method? Why was this journal chosen?]

Response 4: [Thank you for this insightful question.(§3.11) (§3.2.8) : A technology in the field of image encryption, structured around a dynamic library (LIEM) that relies on lightweight registers (two 1-byte registers) for low-footprint management, favoring parallel encryption. This technological approach ensures enhanced security and is scalable against modern attacks, while maintaining low latency suitable for IoT, medical, and cloud applications.

]

Comments 5:[Where are the areas of application of your method?]

Response 5: [

Thank you for this constructive comment. We agree that the application areas of our proposed method should be made clearer. While we briefly mentioned some domains in the conclusion, we have now expanded the discussion to explicitly highlight the main real-world contexts where the DCC can be effectively applied (see the end of section 4).]

Comments 6:[In Table 6, there is bad text in the last line.]

Response 6: [We thank the reviewer for noticing this formatting error. The issue in Table 6 has been corrected.]

Reviewer 2 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

1.The pseudocode format is still not very standardized enough and needs to be revised.

2.The specific definitions of NPCR and UACI should be provided.

3.The number of test images is relatively limited. For both color and grayscale images, I believe that commonly used resolutions such as 256×256, 512×512, and 1024×1024 should be included in the experiments.

4.Recently, multi-image encryption algorithms[Batch image encryption using cross image permutation and diffusion], as well as encryption schemes that integrate parallel computing [A parallel image encryption algorithm using intra bitplane scrambling] and semantic understanding [Semantically enhanced selective image encryption scheme with parallel computing], have been proposed. It is recommended to discuss these algorithms in the Future Work section and elaborate on the possibility of extending the encryption engine to accommodate more such advanced methods in the future.

5. Some comparison results should be reported.

Author Response

Comments 1 :[ The pseudocode format is still not very standardized enough and needs to be revised.]

Response 1 :[We thank the reviewer for this helpful remark. We acknowledge that the pseudocode in the original submission was not fully consistent with standard conventions, which may have affected readability.

]

Comments 2 :[ The specific definitions of NPCR and UACI should be provided]

Response 2 :[Thank you for this valuable comment. We agree that the description of the NPCR (%) and UACI (%) metrics was not sufficiently clear in the original manuscript . We have therefore revised the text to include a more detailed explanation of their meaning, importance in image encryption evaluation, and the mathematical formulas used for their computation.

]

Comments 3 :[ The number of test images is relatively limited. For both color and grayscale images, I believe that commonly used resolutions such as 256×256, 512×512, and 1024×1024 should be included in the experiments.]

Response 3 :[ We thank the reviewer for pointing out this limitation. We agree that testing across multiple resolutions is necessary to validate scalability and robustness.In the revised manuscript, we have expanded the experimental evaluation to include additional test images ]

Comments 4 :[ Recently, multi-image encryption algorithms[Batch image encryption using cross image permutation and diffusion], as well as encryption schemes that integrate parallel computing [A parallel image encryption algorithm using intra bitplane scrambling] and semantic understanding [Semantically enhanced selective image encryption scheme with parallel computing], have been proposed. It is recommended to discuss these algorithms in the Future Work section and elaborate on the possibility of extending the encryption engine to accommodate more such advanced methods in the future.]

Response 4 :[We thank the reviewer for the insightful suggestion. We have updated the Future Work section to discuss recent advances in multi-image, parallelized, and semantic-aware encryption, and we elaborated on how the proposed DCC framework could be extended to accommodate these methods in future research.]

Comments 5 :[ Some comparison results should be reported.]

Response 5 :[ We appreciate the reviewer’s observation. We have expanded Section Comparative Evaluation to include comparison results with recent state-of-the-art schemes, such as the 6D multistable memristive chaotic system and the dynamic Hill cipher for medical IoT. Reported entropy, NPCR, and UACI values from these works were discussed against our proposed DCC results. The comparison highlights that while existing methods achieve strong security metrics, our DCC provides comparable or superior performance with the added advantages of modular adaptability and lightweight implementation.

]

 

Reviewer 3 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Based on the multiple algorithms in the Library of Image Encryption Methods (LIEM), this paper proposes a new Dynamic Cipher Composition (DCC). Simulation analysis was conducted on the algorithm, which proved its resistance to statistical attacks, differential attacks and structural attacks. The article involves a certain amount of work. Some issues should be noted as follows:

1) In the introduction section, the author should review the latest published relevant literature and include relevant literature review sentences.
2) Figure 1 exceeds the page width.
3) The flowchart in Figure 5 is not very good-looking. It is suggested to redraw it and add some photos of the encryption and decryption pictures.
4) For the section on encryption effect evaluation, it is suggested to include some comparisons of image encryption references. Such as Dynamic Analysis and Application of 6D Multistable Memristive Chaotic System with Wide Range of Hyperchaotic States and A novel dynamic hill cipher and its applications on medical IoT
5) The author's contributions, data validity, conflict of interest relationships, etc. should be declared.

Author Response

Comments 1: [In the introduction section, the author should review the latest published relevant literature and include relevant literature review sentences.]

Response 1: [ We thank the reviewer for this valuable suggestion. In the revised manuscript, we have updated the Introduction section to incorporate a more comprehensive review of the most recent and relevant literature. Specifically, we added a discussion of dynamic and multi-algorithm encryption schemes (see section 2)]

Comments 2: [Figure 1 exceeds the page width.]

Response 2: [We thank the reviewer for the suggestion. We have updated the figure]

Comments 3: [The flowchart in Figure 5 is not very good-looking. It is suggested to redraw it and add some photos of the encryption and decryption pictures.]

Response 3: [We thank the reviewer for the suggestion. We have updated the figure 5

]

Comments 4: [For the section on encryption effect evaluation, it is suggested to include some comparisons of image encryption references. Such as Dynamic Analysis and Application of 6D Multistable Memristive Chaotic System with Wide Range of Hyperchaotic States and A novel dynamic hill cipher and its applications on medical IoT]

Response 4: [We thank the reviewer for the suggestion. We have updated the comparative evaluation section to include recent works on the 6D multistable memristive chaotic system and the dynamic Hill cipher for medical IoT. The revised text discusses their performance and highlights that our proposed DCC achieves comparable or superior statistical results while offering higher adaptability and lightweight implementation.]

Comments 5: [ The author's contributions, data validity, conflict of interest relationships, etc. should be declared.]

Response 5: [ Thank you for pointing this out. In the revised version, we have included explicit statements regarding the authors’ contributions, data validity, and conflict of interest:

• Authors’ Contributions: All authors contributed significantly to this work. Specifically, S.D. conceptualized the study and supervised the project; F.G. and J.B. developed the encryption framework and performed the experiments; M.C. contributed to data analysis and validation; M.E.K. provided critical revisions and technical guidance. All authors reviewed and approved the final manuscript.
• Data Validity: The experimental results presented in this paper were generated using standard benchmark images (Lena, Peppers, Baboon, Cameraman) and widely adopted cryptographic evaluation metrics (Entropy, NPCR, UACI, Correlation Coefficients). All experiments were performed in MATLAB R2023b on a controlled system, and the findings are consistent with established benchmarks in the literature.
• Conflict of Interest: The authors declare that there are no conflicts of interest related to this work.

]

Reviewer 4 Report (New Reviewer)

Comments and Suggestions for Authors

Overall Recommendation: Major Revision
1. Weak Literature Review
-The literature review is superficial and lacks critical engagement with existing works.
-References are listed but not synthesised to position the paper’s contribution within the field clearly.
-No comparison with recent state-of-the-art dynamic or multi-algorithm encryption schemes is provided.

2. Methodological Weaknesses
-The selection of encryption methods (AES, Hill, RC4, etc.) is not justified—no rationale is given for why these specific methods were chosen or how they complement each other.
-The combinatorial analysis (M=7, N=3) is presented without sufficient mathematical or security justification.
-The radial segmentation method is described algorithmically but lacks a clear security or performance rationale compared to other segmentation strategies.
-The register-based indexing mechanism is innovative but not sufficiently analysed for potential vulnerabilities (brute-force attacks on 1-byte keys).

3. Limited Technical Analysis and Reflection
-No statistical tests beyond basic metrics (entropy, NPCR, UACI, correlation) are provided. More advanced analyses (key sensitivity, noise resistance, known/chosen-plaintext attacks) are missing.
-The comparative evaluation is weak—only compared to mono-method encryption (AES, Hill), not to other modern dynamic or hybrid encryption schemes.
-Execution time results are presented but not contextualised—no comparison with existing efficient encryption methods to demonstrate practical advantage.
-The discussion section is descriptive rather than analytical—it does not critically reflect on limitations, trade-offs, or scenarios where the method may underperform.

4. Additional Minor Points
-Figures and tables are poorly integrated and sometimes unclear (Figure 1 lacks axis labels and proper captioning).
-The writing contains grammatical errors and awkward phrasing that affect clarity.
-The conclusion is overly optimistic and does not acknowledge the preliminary nature of the results or the need for further validation.

Suggested Improvements:
-Expand the literature review to include a comparative table of recent multi-method encryption techniques.
-Justify method selection and segmentation strategy with references or empirical rationale.
-Include deeper security analysis (key space, sensitivity, attack resilience).
-Compare results with at least 2–3 contemporary encryption schemes.
-Revise figures and improve writing clarity.

Author Response

Comments 1: [ Weak Literature Review]

Response 1: [We appreciate the reviewer’s constructive feedback. The literature review section has been substantially revised to provide a more critical synthesis of existing works. We have added recent state-of-the-art references on dynamic and multi-algorithm image encryption, highlighting their strengths and limitations, and positioned our proposed DCC framework in relation to them. This revision clarifies the novelty and contribution of our work within the field (see introduction section and related work).

]

Comments 2: [ Methodological Weaknesses]

Response 2: [

-We thank the reviewer for all these important observation (§3.1.1) These methods are for illustrative purposes. In an operational deployment, we recommend replacing them with robust, standardized methods

(§5) We have proposed an approach to encryption that supports all methods. The chosen methods were selected to pedagogically illustrate the operation, but they can be replaced by more robust methods without any change to the method architecture.

-in section (§2.3) of the revised article, we have discussed and explained the  mathematical and security justification.

-In (§2.5) of the revised article, we justify that radial segmentation increases inter-region entropy by breaking linear spatial structures; reserving visual coherence for sensitive applications (e.g., medical imaging).

In the revised article, to answer this question, to fill this gap, we have added a new paragraph.

-The new subsection « 2.7 Register Security Assessment and Scalability Strategy »  which offers extensibility of the key space without any change to the system architecture

]

Comments 3: [  Limited Technical Analysis and Reflection]

Response 3: [We thank the reviewer for this valuable suggestion.

- We have improved the article by introducing an experimental study evaluated the method with the most standard metrics, namely:

- (§ 3.2.4) Histogram Analysis

- (§ 3.2.7). Correlation Analysis

- (§ 3.2.4 to 3.2.7): Entropy, NPCR, UACI, and correlation coefficients are analyzed.

- (§ 3.2.1): Key Sensitivity

- (§ 3.2.2): Noise Resistance

- (§ 3.2.3): Resistance to Known/Chosen-Plaintext Attacks

NIST tests are planned for future work, adopting new, more robust methods for the framework.

 

      - A benchmarking has been added in ta (Table 1)  , and a comparison has been introduced in (§3.3)

-

-Section (§3.2.8): "Execution Time and Efficiency" gives the encryption/decryption times for 4 images (e.g., Lena: 84.1 ms, Mandrill: 44.7 ms), which present very low and acceptable execution times. And in future work, we intend to address complexity instead of execution time.

- In the revised version, the Discussion section serves as a summary of achievements rather than a critical analysis. It also explores the potential applications of this method in real contexts.

 ]

Comments 4: [Additional Minor Points ]

Response 4: [ We appreciate the reviewer’s attention to presentation quality.

-In this revised version, we have enhanced the integration of figures, tables, and other visual elements.

- In (§2)a  new section has been added to provide a literature review covering recent work in the field. with a detailed table for comparison

-The justification of the segmentation strategy is mainly addressed in (§2.5) of the article, relying on qualitative arguments and external references. A theorem with a comprehensive analytical and empirical demonstration is possible in future work.

-We have improved the article by introducing an experimental study evaluated the method with the most standard metrics, namely:

- (§ 3.2.4) Histogram Analysis

- (§ 3.2.7). Correlation Analysis

- (§ 3.2.4 to 3.2.7): Entropy, NPCR, UACI, and correlation coefficients are analyzed.

- (§ 3.2.1): Key Sensitivity

- (§ 3.2.2): Noise Resistance

- (§ 3.2.3): Resistance to Known/Chosen-Plaintext Attacks

NIST tests are planned for future work, adopting new, more robust methods for the framework.

 

-A comparison has been introduced in (§3.3) , and a benchmarking has been added in ta (Table 01) 

-This revised version has been prepared with full consideration of your valuable observations.]

 

Round 2

Reviewer 2 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

1.What is the theoretical innovation of this method?

2.The title of this paper remains unclear and overly broad. What does "a set of methods" refer to? The title needs to be revised, and it can be refined by incorporating the core principles of these methods.

3.In Sec.6, the references related to the newly added content should be included in the reference list.

Author Response

Comments 1:[What is the theoretical innovation of this method? ]

Response 1: [The theoretical innovation of our method lies in the introduction of a DCC framework that formalizes encryption as a combinatorial and register-based process rather than relying on a single algorithmic structure. Specifically, the approach combines three novel principles: (1) randomized multi-method selection from a library of image encryption methods (LIEM), mathematically modeled to balance diversity and efficiency under lightweight indexing constraints; (2) region-wise heterogeneous encryption through structured segmentation, ensuring ciphertext variability across both space and rounds; and (3) a dual-register configuration system that compactly encodes method combinations and their application modalities, enabling tens of thousands of admissible configurations with minimal overhead. Together, these principles establish a new theoretical paradigm in image encryption, where cryptographic strength emerges from controlled randomness, spatial diversity, and efficient register-based indexing, achieving both adaptability and lightweight performance not attained by existing schemes.]

Comments 2:[The title of this paper remains unclear and overly broad. What does "a set of methods" refer to? The title needs to be revised, and it can be refined by incorporating the core principles of these methods. ]

Response 2: [We sincerely thank the reviewer for this valuable observation. We agree that our previous title was somewhat broad and could lead to ambiguity regarding the term “a set of methods.” To address this, we have revised the title to more clearly reflect the core principles and contributions of our work. Specifically, the new title highlights both the  DCC approach and its reliance on a LIEM combined with spatial segmentation and lightweight register-based management.

The revised title is: « Image-Type Data Security via Dynamic Cipher Composition from Method Libraries »]

Comments 3:[In Sec.6, the references related to the newly added content should be included in the reference list. ]

Response 3: [We thank the reviewer for this observation. In Section 6, we had indeed added new content related to future research directions. To address this comment, we have carefully revised the section and integrated the corresponding references into the reference list, ensuring that all newly introduced concepts are properly supported by recent and relevant literature.]

Reviewer 3 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Accept in present form

Author Response

We sincerely thank you for your valuable feedback and for recommending the acceptance of our article. We truly appreciate the time and effort you devoted to reviewing our work

Reviewer 4 Report (New Reviewer)

Comments and Suggestions for Authors

Most of my round 1 remarks and recommendations were addressed - recommend acceptance

Author Response

We are grateful for your thoughtful review and for accepting our article. Your comments and feedback have been very helpful in improving the quality of our work, and we deeply appreciate your support

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This work proposes a new image encryption architecture, which deviates from the traditional single-method encryption. Practical experimental studies were conducted using classic methods such as AES, Hill, RC4, S-box, SPN, DNA encoding and chaotic permutation. The results demonstrated the effectiveness of this method. This article seems to be a bit lacking in terms of novelty and workload.

1）There are problems with the format norms of the article, including the use of grammar and the format norms of references, etc.
2）Algorithm 1 and Algorithm 2 should be written in the format table of pseudo-code.
3) The flowchart in Figure 5 needs to be redrawn.
4) In the image encryption section, all are common encryption methods. No new performance analysis was seen, nor was there any comparison with the latest published references.

Comments on the Quality of English Language

There are problems with the format norms of the article, including the use of grammar and the format norms of references, etc.

Author Response

Comments1: [There are problems with the format norms of the article, including the use of grammar and the format norms of references, etc.]

Response1: We appreciate this observation. We have carefully revised the manuscript to correct grammatical inconsistencies, improve sentence clarity, and ensure smooth flow throughout. Additionally, we have reformatted the references according to the journal’s guidelines, ensuring consistent citation style and completeness.

Comments2: [Algorithm 1 and Algorithm 2 should be written in the format table of pseudo-code.]

Response2: Thank you for this suggestion. We have rewritten Algorithm 1 and Algorithm 2 using a structured pseudo-code table format with clear indentation, input/output specifications, and stepwise operations to enhance readability and align with standard pseudo-code presentation conventions.

Comments3: [The flowchart in Figure 5 needs to be redrawn.]

Response3:  We agree with this observation. Figure 5 has been redrawn with clearer layout, consistent font, and improved visual structure to enhance readability and accurately represent the encryption process steps. We believe this will help readers follow the proposed engine workflow more intuitively.

Comments4: [In the image encryption section, all are common encryption methods. No new performance analysis was seen, nor was there any comparison with the latest published references.]

Response4: Thank you for this critical point. While our proposed architecture uses commonly known encryption methods, its novelty lies in dynamic multi-method selection, region-specific encryption, and register-based modality assignment, which are not standard in classical single-method encryption systems.

 

Reviewer 2 Report

Comments and Suggestions for Authors

1. Have you used software to assess encryption strength? Any encryption method involves programs that test for resistance to cracking. Provide this.
2. Do you use different types of segmentation? Which type of segmentation is the most robust?
3. There is no discussion section and no further work directions.
4. Where can this method be applied?
5. Should Algorithm 1 be presented in this form? This is unacceptable. It is better to provide a flowchart.
6. Table 4 contains text that overlaps the table. Please fix this.

Author Response

Comments1:[Have you used software to assess encryption strength? Any encryption method involves programs that test for resistance to cracking. Provide this.]

Response1:

We thank the reviewer for raising this important point. In our study, the encryption strength of the proposed engine was rigorously evaluated using standard cryptographic and statistical analysis tools that are widely employed in image encryption research. Specifically, we assessed resistance to cracking and robustness against common cryptanalytic attacks using the following approaches:

1. Statistical Tests – We implemented Shannon entropy analysis, histogram uniformity, and adjacent pixel correlation tests. The encrypted images consistently achieved entropy values close to 8 (ideal randomness) and near-zero correlation coefficients, demonstrating resistance to statistical attacks.
2. Differential Attack Analysis – We used MATLAB-based programs to compute NPCR (Net Pixel Change Rate) and UACI (Unified Average Changing Intensity) by altering one pixel in the plaintext image. Results showed NPCR > 99.6% and UACI ≈ 33%, which exceed the commonly accepted thresholds for strong resistance to differential attacks.
3. Comparative Software-Based Evaluation – We compared our results against baseline mono-method encryptions (AES-only, Hill-only) and against benchmarks reported in recent literature. Our engine consistently outperformed these methods in randomness and sensitivity metrics.
4. Execution Environment – All encryption and testing programs were developed and executed in MATLAB R2023b on an Intel Core i7-12700 system with 32 GB RAM. These software-based evaluations replicate standard cryptanalytic testbeds used in the field and ensure reproducibility of results.

Comments2:[ Do you use different types of segmentation? Which type of segmentation is the most robust?]

Response2:

We thank the reviewer for this insightful question. In our proposed engine, we designed the framework to support multiple segmentation strategies for dividing the image into disjoint regions before encryption. As explained in Section 2.3, we considered both linear segmentation (uniform rectangular/grid-based partitions) and non-linear segmentation (e.g., radial or angular divisions).

For the experimental validation presented in this work, we selected radial segmentation because:

• It disrupts structural patterns more effectively than simple grid partitioning.
• It increases inter-region entropy by reducing residual correlations between adjacent regions.
• It provides enhanced robustness against structural and statistical attacks, as the wedge-shaped regions prevent attackers from exploiting repetitive rectangular structures in ciphertexts.

Comments3:[There is no discussion section and no further work directions.]

Response3: We thank the reviewer for pointing this out. In the revised version, we have explicitly added Sections 4 (Discussion) and 5 (Future Work). Although these sections were already present in the original submission, we recognize that their placement and level of detail may not have highlighted them clearly as standalone contributions..

Comments4:[Where can this method be applied?]

Response4: We thank the reviewer for raising this important point. The proposed engine is designed as a general-purpose image encryption framework, and its modular, lightweight architecture makes it suitable for several real-world application domains where visual data security is critical.  

We have now included these application domains in the revised Conclusion section to explicitly highlight the practical scope of our method.

Comments5:[Should Algorithm 1 be presented in this form? This is unacceptable. It is better to provide a flowchart.]

Response5: We appreciate the reviewer’s suggestion regarding the presentation of Algorithm 1. In the revised manuscript, we have replaced the original algorithmic representation with a table to enhance clarity and improve readability. We believe this new format makes the process more intuitive for the readers.

Comments6:[Table 4 contains text that overlaps the table. Please fix this.]

Response6:

Thank you for pointing this out. We have corrected the formatting issue in Table 4 to ensure that all text is clearly visible and does not overlap with the table boundaries. We have also adjusted the column widths and font size to improve readability.

Reviewer 3 Report

Comments and Suggestions for Authors

1.The pseudocode format is not very standardized and needs to be revised.

2.The authors should clearly specify the inputs and outputs of each algorithm.

3.The specific definitions of NPCR and UACI should be provided.

4.The number of test images is relatively limited. For both color and grayscale images, I believe that commonly used resolutions such as 256×256, 512×512, and 1024×1024 should be included in the experiments.

5.Recently, multi-image encryption algorithms[Batch image encryption using cross image permutation and diffusion], as well as encryption schemes that integrate parallel computing [A parallel image encryption algorithm using intra bitplane scrambling] and semantic understanding [Semantically enhanced selective image encryption scheme with parallel computing], have been proposed. It is recommended to discuss these algorithms in the Future Work section and elaborate on the possibility of extending the encryption engine to accommodate more such advanced methods in the future.

Author Response

Comments1: [The pseudocode format is not very standardized and needs to be revised.]

Response1:

We thank the reviewer for pointing this out. In the initial version, the pseudocode was written in a descriptive step-by-step manner, which may not have fully aligned with standardized pseudocode presentation. To address this, we have revised all algorithms (Register1 generator, segmentation, and Register2 applicator) into a uniform and more readable pseudocode style that follows common conventions (structured indentation, “Input/Output” declaration, clear control structures such as For, If–Else, While).

Comments2:[The authors should clearly specify the inputs and outputs of each algorithm.]

Response2: 

We thank the reviewer for this valuable suggestion. In the original version, the algorithms were described with steps, but the inputs and outputs were not always explicitly defined in a consistent manner. To improve clarity, we have revised all pseudocode listings to include a standardized “Input:” and “Output:” header at the beginning of each algorithm.

Comments3:[The specific definitions of NPCR and UACI should be provided.]

Response3:We thank the reviewer for this valuable comment. In the revised manuscript, we have included the formal definitions of NPCR (Number of Pixels Change Rate) and UACI (Unified Average Changing Intensity) in Section 3.2.

Comments4:[The number of test images is relatively limited. For both color and grayscale images, I believe that commonly used resolutions such as 256×256, 512×512, and 1024×1024 should be included in the experiments.]

Response4:We thank the reviewer for raising this important point. While we agree that evaluating images with multiple resolutions (256×256, 512×512, 1024×1024) would provide additional validation, our current experimental design focused on standard benchmark images (Lena, Peppers, Baboon, Cameraman) that are widely adopted in the image encryption literature and cover both grayscale and color formats. These images, particularly at resolutions of 256×256 and 512×512, are commonly used as reference cases and provide a fair basis for comparative analysis with existing works. Furthermore, the proposed Dynamic Cipher Composition (DCC) is resolution-independent, as its segmentation and encryption mechanisms operate generically on any image size. This ensures that the method can be directly applied to higher resolutions without modification. We will highlight this aspect more clearly in the revised manuscript.

Comments5:[Recently, multi-image encryption algorithms[Batch image encryption using cross image permutation and diffusion], as well as encryption schemes that integrate parallel computing [A parallel image encryption algorithm using intra bitplane scrambling] and semantic understanding [Semantically enhanced selective image encryption scheme with parallel computing], have been proposed. It is recommended to discuss these algorithms in the Future Work section and elaborate on the possibility of extending the encryption engine to accommodate more such advanced methods in the future.]

Response5:We thank the reviewer for this valuable suggestion. In the revised manuscript, we have expanded the Future Work section to include a discussion of recent advanced image encryption approaches, including multi-image encryption, parallel computing-based schemes, and semantically enhanced methods. Furthermore, we have emphasized the potential of extending our encryption engine to integrate such techniques in future research.