Enhanced Copy-Move Forgery Localization Identification: Leveraging Features for Optimized Detection
Main Article Content
Abstract
- Image forgery detection has become significantly important in the digital era, as the genuineness of visual output is frequently compromised. This study addresses the growing need for robust techniques to detect image forgeries, particularly copy-move forgeries, which are common and difficult to detect due to the sophisticated methods used by forgers. The study presents an improved ensemble model for detecting and locating copy-move fraud using a powerful combination of machine learning and neural networks. Convolutional Neural Network (CNN) is used for extracting features and capturing complex patterns and details in pictures, while XGBoost is used for classification, taking advantage of its great efficiency and accuracy in processing big datasets. The proposed ensemble model successfully detects and accurately pinpoints forgeries in digital photographs and surpasses the performance of current approaches on the MICC-F600 and MICC-F2000 datasets, attaining F1 scores of 98.59 and 99.03, respectively. Additionally, the ensemble model achieves accuracy, precision, and recall rates of 99%, 98.66%, and 98.62% on the MICC-F600 dataset, and 99%, 98.5%, and 98.03% on the MICC-F2000 dataset. The results clearly indicate the method's exceptional accuracy in detecting copy-move forgeries, establishing it as a dependable tool for digital forensics. The experimental results highlight the method's capacity for practical applications in detecting picture counterfeiting, providing a substantial enhancement compared to conventional approaches.
Article Details

This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.