Subsampling-based robust watermarking using neural network detector

This paper presents a robust digital image watermarking scheme by using neural network detector. First, the original image is divided into four subimages by using subsampling. Then, a random binary watermark sequence is embedded into DCT domain of these subimages. A fixed binary sequence is added to the head of the payload watermark as the samples to train the neural network detector. Because of the good adaptive and learning abilities, the neural network detector can nearly exactly extract the payload watermark. Experimental results show good performance of the proposed scheme on resisting common signal processing attacks.