Detection of Tampering by Image Resizing Using Local Tchebichef Moments

Round 1

Reviewer 1 Report

The paper study the detection of image resizing using local Tchebichef moments. A series of experiments have been completed. It can be come to conclusion from the experimental results that the presented method can get better accuracy ratio. Nevertheles, the presented approach is verified the performance for the image resizing with high scaling ratios.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The title sounds: Detection of image resizing using local Tchebichef
moments

however, in the abstract, the description indicates that the article concerns the use of LTM for image scaling:
"Therefore, a LTM-based universal detection approach is proposed for image resizing". It is confusing.

The reviewer assumes that the authors were concerned about the use of the LTM method in the task of detecting manipulation on images by scaling.

By the way reviewer suggests clarifying the title:
"Detection of tampering by image resizing using local Tchebichef moments".

It is recommended to edit the abstract so that it accurately describes the purpose and tasks of the article.

It is recommended to submit the article for correction by a native speaker. The current grammatical structure of the article (especially the abstract, the introduction, the second chapter and the conclusion) makes it difficult to understand.

Image retargeting includes:
resizing, recompositing, reshuffling, inpainting.
Does the article only concern the detection of changes in size? Yes, and it would be good to mention it in the abstract and the introduction.

In the case of Figure 4, it is recommended to change the colors to black (or shades of gray) on a white background. Currently, graphics on figures 4.d) e) f) are unreadable.

In chapter 4, the sentence should be slightly corrected:
It can be found from the table 3 that the presented method can get best accuracy
among the three content-aware resizing methods mentioned in this paper. However, the proposed
method can’t obtain a good detection accuracy on the tampered images with JPEG compression,
moreover, the false positive rate is relatively high for seam carving method and scaling method.

in the form of:
It can be found from the table 3 that the presented method can get high accuracy for the three content-aware resizing methods mentioned in this paper. However, the proposed method can’t obtain a good detection accuracy on the tampered images with JPEG compression.
Moreover, the false positive rate is relatively high for seam carving (SC) method and scaling method (SL).

In the tables 3 and 5 would be good to write "detection accuracy" instead "classified accuracy".
In the titles of the tables 2-3 and 4-5 would be good to add which results are obtained "with preprocessing" and which are "without preprocessing".

Author Response

Please see the attachment.

Author Response File: Author Response.pdf