Copy-move forgery detection using ASIFT

Karsh, Ram Kumar; Das, Akhil; Swetha, G.; Medhi, Abhishek; Laskar, Rabul Hussain; Arya, Utkarsh; Agarwal, Rohit Kumar

doi:10.1109/iicip.2016.7975329

Cited by 5 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thereafter, Huang et al [80] and Ardizzone et al [65] extracted feature as SIFT descriptor [80], and the search method best-bin-first is used to suit for same type of features. Li et al [81] used the algorithm of expectation maximization and Dense Scale-Invariant Feature Transform (DSIFT) to evaluate the transformation matrix whereas Karsh et al [82] uses Affine-Scale-Invariant Feature Transform (ASIFT) that identifies copy-move tampering. Typical keypoint-based strategies are SIFT, ASIFT, DSIFT, MIFT i.e.…”

Section: Scale Invariant Feature Transformationmentioning

confidence: 99%

Comparative Analysis of Techniques Used to Detect Copy-Move Tampering for Real-World Electronic Images

Tripathi,

Kumar,

Tripathi

2023

IJIE

View full text Add to dashboard Cite

Evolution of high computational powerful computers, easy availability of several innovative editing software package and high-definition quality-based image capturing tools follows to effortless result in producing image forgery. Though, threats for security and misinterpretation of digital images and scenes have been observed to be happened since a long period and also a lot of research has been established in developing diverse techniques to authenticate the digital images. On the contrary, the research in this region is not limited to checking the validity of digital photos but also to exploring the specific signs of distortion or forgery. This analysis would not require additional prior information of intrinsic content of corresponding digital image or prior embedding of watermarks. In this paper, recent growth in the area of copy-move tampering identification have been discussed along with benchmarking study. In this study, three combination of methods like discrete cosine transform with Fast K-Means, discrete wavelet transform with Fast K-Means and discrete cosine transform with Fuzzy-C means can efficiently locate duplicated regions with quantitative parameter i.e. 92%, 83% and 92% respectively. With variety of methodologies and concepts, different applications of forgery detection have been discussed with corresponding outcomes especially using machine and deep learning methods in order to develop efficient automated forgery detection system. The future applications and development of advanced soft-computing based techniques in digital image forgery tampering has been discussed.

show abstract

Section: Scale Invariant Feature Transformationmentioning

confidence: 99%

Comparative Analysis of Techniques Used to Detect Copy-Move Tampering for Real-World Electronic Images

Tripathi,

Kumar,

Tripathi

2023

IJIE

View full text Add to dashboard Cite

show abstract

“…ASIFT simulates different image views by varying the orientation of the camera axis parameters latitude and longitude. Some papers [26, 27] applied ASIFT for copy–paste forgery detection. Though ASIFT is claimed to be fully affine, it is not scaled invariant as that of SIFT.…”

Section: Related Workmentioning

confidence: 99%