Invisible Watermark Embedding and Retrieval for Facial Image Copyright Protection and Forensics

Ensuring digital image security for facial images, has become increasingly important. This demand calls for robust watermarking techniques to protect images copyright. This paper will dwell on a design targeting a web application for secure watermark embedding and retrieval, with emphasis on facial image watermarking. It makes use of machine learning for face detection and techniques of encryption to protect the watermark that will be embedded. Benchmarking on robustness, security, and performance turns out incredibly improved with respect to traditional approaches. The main contributions presented in this paper are a user-friendly interface, safe and efficient face detection for watermarking purposes, and very good security measures to protect not only the watermark but also the host image.