Paper Contents
Abstract
These days advanced photo editing tools make it very easy to create fake or modified digital images. This has become a serious problem because its hard to tell whether an image is real or not, and it affects the trust and originality of images. Since many image manipulations are too complex to be noticed by the human eye, there is a strong need for system that can detect image forgery. Most studies in this area use passive detection methods, which means they check the image itself for signs of editing without using extra information like watermarks or digital signatures. Nowadays, deep learning (DL), especially Convolutional Neural Networks (CNNs), has been used to improve the accuracy of forgery detection. Many researchers use transfer learning, by using pre-trained models like AlexNet, VGG16, MobileNetV2, and ResNet50v2 for classifying and locating forgeries. Some methods combine Error Level Analysis (ELA) with CNNs to spot differences. Others use Graph Convolution Networks (GCNs) to understand spatial patterns and relationships in images. This review paper highlights the different types of forgeries and the techniques used to detect them and also future research directions in image forgery detection.
Copyright
Copyright © 2025 PRAJKTA RAJESH GIRHE. This is an open access article distributed under the Creative Commons Attribution License.
