Design and Implementation of Image Forgery Detection System Based on Cloud Computing

Abstract

The masses of people urgently need a platform to help them identify the authenticity of network images, but the existing image forgery detection algorithms generally have high thresholds and poor real-time problems, making it difficult for the public to provide efficient detection services. Therefore, this paper proposes a cloud-based network image forgery detection system that uses B/S architecture, integrates multiple algorithms, and uses cloud computing and GPU computing technologies to improve system throughput and detection speed. Experimental results show that the system can effectively detect copy-move and splicing forgeries, and the GPU's speedup ratio reaches 6.5, which significantly improves the detection speed and throughput of the system.