Image Content Forgery Detection Model Combining PSO and SVM in Electronic Data Forensics

Abstract

With the rapid progress of information technology, the Internet is saturated with copious amounts of data and visuals. However, with the widespread availability of different image editing software, counterfeit image content arises periodically. To tackle image content forgery, the research is founded on the Gaussian mixture distribution similarity measure image forgery detection algorithm. The image classifier underwent training through encoding its underlying features and utilizing the encoded data as inputs for the Support Vector Machine. Optimization of the Support Vector Machine was performed simultaneously using the Improved Particle Swarm Optimization algorithm. The results indicate that the SVM-based image content forgery detection model, which employed improved Particle Swarm Optimization, achieved a detection rate of 94.89% and processed the images in 22.06 milliseconds. In summary, the study of an image content forgery detection model that combines improved Particle Swarm Optimization and Support Vector Machine in electronic data forensics has resulted in a high detection accuracy.