I used to saw a lot of encryption papers that have small significance in security improvements especially in terms of entropy, histogram and coefficient correlation especially in image encryption. Why is this happening?i mean why still publish a proposed method that have 0.001 or so on improvement in terms of entropy or coefficient correlation?even some of the papers does not tell how it significantly differs in terms of security.

for example in this paper, in terms of entropy and npcr uaci it does not have significant improvement compared to the referenced paper.. like in entropy value it only have 0.001 difference to ref 33 and 34 and also in npcr uaci only 0.01 difference.. i know its value is close to the ideal value hence it can resist entropy attack and so on, but why make another proposed method with no significant contribution in the security?is it just for encryption method variation?i am new in cryptography, please help enlighten me

  • 4
    $\begingroup$ It might be useful to name or link some of the papers you reference, and maybe point out what difference exactly you mean. $\endgroup$ – tylo Dec 26 '20 at 14:09
  • $\begingroup$ @kelalaka thank you $\endgroup$ – Muhammad Fauzi Dec 26 '20 at 18:25
  • 1
    $\begingroup$ @tylo i have updated the question $\endgroup$ – Muhammad Fauzi Dec 26 '20 at 18:26
  • $\begingroup$ Because we are not like in the older times, the paper count! $\endgroup$ – kelalaka Dec 26 '20 at 19:08
  • 3
    $\begingroup$ Nowadays image is compressed, and then becomes arbitrary data, which can be compressed using any general purpose algorithm, such as AES-GCM. So there is nothing special about image encryption. Every paper on image encryption is entirely pointless from the standpoint of its contribution to cryptography. Why so many papers on image encryption are published has nothing to do with cryptography, thus is off-topic. $\endgroup$ – fgrieu Dec 26 '20 at 21:21