Current Result Document :
| 한글제목(Korean Title) |
Exploiting Multiple Color Models to Investigate Photo Forgery |
| 영문제목(English Title) |
Exploiting Multiple Color Models to Investigate Photo Forgery |
| 저자(Author) |
Saurabh Agarwal
Ki-Hyun Jung
|
| 원문수록처(Citation) |
VOL 45 NO. 01 PP. 2148 ~ 2152 (2022. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Digital photo forgery can be accomplished simply by the support of any good quality photo editing software. Most of the time, fake photos are used to disseminate misleading information. So, there is a need for some sort of technique to distinguish between a counterfeit photo and a pristine photo. In this paper, multiple representations of photos in different color models - RGB, YCbCr, and LAB - are considered. Further, Local binary patterns (LBP) and co-occurrence variants as texture descriptors are used to extract features. The texture descriptors give promising results in many applications such as face recognition and biometrics. Texture descriptors can also highlight abrupt changes in the fake photo. The performance is evaluated on CASIA 1.0 and Columbia publicly available databases. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
