Discrete Wavelet Transform Based Multiple Watermarking for Digital Images Using Back-Propagation Neural Network

A Discrete Wavelet Transform (DWT) based multiple watermarking technique for images using Back-Propagation neural networks (BPNN) are proposed. The successive/re- watermarking is one of the best method in multiple watermarking techniques. In successive/re-watermarking method, the various watermarks are embedded and extracted one by one. The wavelet coefficient is selected based on the weight factor by using Human Visual System (HVS). The BPNN is incredibly well-liked in neural networks and its variety of supervised learning neural networks. The two watermarks are embedding into the original image using improved BPNN, which can advance the speed of the erudition, reduce error and the trained neural networks be capable of extracting the two watermarks from the embedded images. The simulation results show that the proposed work achieves good quality on the embedded images and more robustness on extracted two watermarks.