Full-Reference Image Quality Assessment Based on Multi-Channel Visual Information Fusion

Round 1

Reviewer 1 Report

My few comments about this manuscript are listed below.

1. Introduction must be improved. In this section, authors are some how unable to list their contributions.

2. Section 2 is just messed up with equations and the parameters. Reader is some how unable to find the proposed solution in this section.

3. Novelty of this manuscript is questionable. A pseudo code placed in this section would have been helpful for reader.

4. Section 3 presents large number of experiments. However, no discussion is listed here. I would suggest authors to include a critical discussion section about their findings.

5. In Section 1, authors should also include the limitations and strengths of few methods from literature, which they have listed.

6. Authors should also include the limitations of their proposed method.

7. Again, I would say, what is the novelty in this manuscript? On a similar topic, large number of publications are available.

8. Section 4 also needs improvement. Moreover, this section should also be flourished with possible future research direction.

Author Response

All the questions answered by the author have been sorted in the attachment, please download to view，thanks a lot.

Author Response File: Author Response.pdf

Reviewer 2 Report

The topic of this paper is very interesting however there are some concerns should be considered before acceptance:

1. The organization and structure of this paper must be improved signifcantly.

2. Improve quality of figures and English language.

3. Add pseduo code for your proposed model.

4. Add related works section and focus and  recent approaches for image quality assessment (AI based).

5. I have noticed the authors did not compare with recent AI based approaches for image quality assessments, hence the authors have to compare with al least 4 to 5 approaches in this category and report results.

6. The authors have to mention the applications of image quality in the introduction and related works section and I recommend adding the following references for this purpose:

- Quality based approach for adaptive face recognition.

-Biometric templates selection and update using quality measures.

-Image quality guided approach for adaptive modelling of biometric intra-class variations.

7. Add example images from four databases with objective and subjective evaluations (metric values) for proposed model and compared approaches.

 

The quality of English language is low and there is a need to be improved significantly.

Author Response

All the questions answered by the author have been sorted in the attachment, please download to view, thanks a lot.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper proposes a Full-Reference Image Quality Assessment Based on Multi-Channel Visual Information Fusion. The presented The methodology begins with converting red, green, blue (RGB) images into the luminance(L), red‒green opponent color channel (M), blue‒yellow opponent color channel (N) or LMN color space. Subsequently, the LoG filter is separately applied to the L, M, and N color channels. The convoluted components are then fused to generate a contrast similarity map using the root-mean-square method, while the chromaticity similarity map is derived from the color channels. Finally, multichannel LoG filtering, contrast, and chromaticity image features are merged. The standard deviation method is then used for sum pooling to create a full-reference IQA computational model. Overall a number of important details are missing and the novelty and contribution of the paper are quite limited. I wish however the authors consider the following major issues before my recommendation for publication.

1.                I suggest to include the significance of this paper in introduction. Please make it clear and add enough original studies. The authors should explain the idea behind their proposed method. Why do the authors think this method should work?

2.                           How does the proposed full-reference IQA method integrate multi-channel visual information in color images?

4.                What are the color channels used in the LMN color space conversion of RGB images?

5.                How are the convoluted components from the LoG filter fused to generate a contrast similarity map in the proposed method?

6.                What is the approach used to derive the chromaticity similarity map in the proposed model?

7.                How are the multi-channel LoG filtering, contrast, and chromaticity image features merged in the proposed full-reference IQA computational model?

8.                What method is used for sum pooling in the proposed model to create a full-reference IQA score?

9.                How were the distorted images from the four image databases used to validate the proposed model?

10.             What are the Pearson linear correlation coefficient values obtained from the evaluation of the proposed model on the image databases?

11.             How does the proposed model compare to existing methodologies in terms of visual correlation prediction accuracy?

12.             What are the strengths and limitations of the proposed full-reference IQA model based on multi-channel visual information fusion for image quality assessment?

1.                The language usage throughout this paper need to be improved, the author should do thorough checking on grammatical and typographical mistakes. Give the article an effective language revision to get rid of few complex sentences that hinder readability and eradicate typo errors.

Author Response

All the questions answered by the author have been sorted in the attachment, please download to view, thanks a lot.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thanks for addressing all suggestions.

Reviewer 2 Report

No more comments