JJEM: Special Edition 2 (December 2024)
JJEM: Special Edition 2 (December 2024)
2024-12-08
Image Forgery Detection Using Deep Learning.
Poornima C N, Sunitha G P
In the contemporary era, digital images are a vital source of information in the modern day, and as such, they are regularly disseminated on social media. False identification could be inferred from the deceptive information. With the many tools and methods that are available today, anyone can quickly create image forgeries that could pose a number of problems for society. Many of the strategies for identifying false identification have been explored in the literature survey that is currently available, but they are unable to provide an accurate result in real-world scenarios. Instead, they can only identify a single type of falsification in an image, such as cloning or resizing. This paper introduces an AI-driven image tampering identification system that will identify various forms of image manipulation. In order to detect forgeries, this study suggests a convolutional neural network-based model. The data will be collected as images and preprocessed by identifying any redundant information or missing values. The image yields several attributes, including dimensions, hue, length, breadth, and height. CNN receives all of the derived characteristics for training. The model uses the CMF technique to identify forgeries and then classifies the image as either forged or not. If an image is forged, it outputs the image containing the location of the faked image. The recommended method will reveal different images based on actual events and produce a 98% accuracy rate in counterfeit detection.
CNN-Convolutional Neural Network, CMF-Copy-Move Forgery, ML-Machine Learning.