Fuzzyfake: Copyright Protection of Deep Fake Images Using FIS Optimization
摘要
Evolution of digital images in present time has added a new dimension in the area of multimedia communication. There is a progressive technological advancement for secure multimedia content distribution. But this rapid development needs to be protected from intruders. Thus, a new image water-marking technique using fuzzy logic is proposed in this paper. The presented technique uses face detection method in OpenCV using Haar cascade library. The extracted face is used as cover image which is being further processed by discrete wavelet transform (DWT)–singular value decomposition (SVD) for watermarking. Further optimization is done to enhance the effectiveness of the proposed technique. Luminance, edge Sensitivity and roughness parameters are used for the evaluation of optimal scaling factor. Deep fake infringement is protected by this proposed method in an effective way ensuring copyright protection. Thus, the proposed method exhibits better results in comparison to existing techniques.