Welcome to the InfoSci Platform
Reference Hub4
Copy-Move Forgery Detection Based on Automatic Threshold Estimation

Copy-Move Forgery Detection Based on Automatic Threshold Estimation

Aya Hegazi, Ahmed Taha, Mazen Mohamed Selim
Copyright: © 2020 |Volume: 12 |Issue: 1 |Pages: 23
ISSN: 1941-6253|EISSN: 1941-6261|EISBN13: 9781799805670|DOI: 10.4018/IJSKD.2020010101
Cite Article Cite Article

MLA

Hegazi, Aya, et al. "Copy-Move Forgery Detection Based on Automatic Threshold Estimation." IJSKD vol.12, no.1 2020: pp.1-23. http://doi.org/10.4018/IJSKD.2020010101

APA

Hegazi, A., Taha, A., & Selim, M. M. (2020). Copy-Move Forgery Detection Based on Automatic Threshold Estimation. International Journal of Sociotechnology and Knowledge Development (IJSKD), 12(1), 1-23. http://doi.org/10.4018/IJSKD.2020010101

Chicago

Hegazi, Aya, Ahmed Taha, and Mazen Mohamed Selim. "Copy-Move Forgery Detection Based on Automatic Threshold Estimation," International Journal of Sociotechnology and Knowledge Development (IJSKD) 12, no.1: 1-23. http://doi.org/10.4018/IJSKD.2020010101

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Recently, users and news followers across websites face many fabricated images. Moreover, it goes far beyond that to the point of defaming or imprisoning a person. Hence, image authentication has become a significant issue. One of the most common tampering techniques is copy-move. Keypoint-based methods are considered as an effective method for detecting copy-move forgeries. In such methods, the feature extraction process is followed by applying a clustering technique to group spatially close keypoints. Most clustering techniques highly depend on the existence of a specific threshold to terminate the clustering. Determination of the most suitable threshold requires a huge amount of experiments. In this article, a copy-move forgery detection method is proposed. The proposed method is based on automatic estimation of the clustering threshold. The cutoff threshold of hierarchical clustering is estimated automatically based on clustering evaluation measures. Experimental results tested on various datasets show that the proposed method outperforms other relevant state-of-the-art methods.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.