Due to availability of the images through internet and many digital media, there is a serious threat to them from the digital thieves. As a consequence, the ownership of the image might be misinterpreted. In this context, research work is needed to resolve rightful ownership. The owner should be able to hide some information in the image and when needed he/she should be able to extract that information to prove his ownership. There has been significant research work in to Digital Water Marking (DWM). The process of embedding information into multimedia object can be termed as watermarking. Also referred to as simply watermarking, a pattern of bits inserted into a digital image, audio or video file that identifies the file's copyright information.
The proposed system embeds watermark by decomposing the host image. Dividing these coefficients into small blocks, calculating the standard deviations of these blocks, deciding whether this block can be use for embedding watermark. The watermark bits are added to the selected coefficient block without any perceptual degradation for host image .The watermark used for embedding is a binary logo image, which is very small compared to the size of the host image. During the watermark recovery, trained probabilistic neural network is employed to extract the watermark.
To ensure the watermark safety and imperceptibly, embedding the watermark bits into the edges and textures of the image we make use of the statistical properties of the dual-tree wavelet transform (DTCWT) and the human visual system (HVS). Due to neural network possessing the learning capability from given training patterns, this method can memorize the relations between a watermark and the corresponding watermarked image.
This algorithm is proposed by Xian-Bin Wen • Hua Zhang • Xue-Quan Xu • Jin-Juan Quan (Published online: 7 June 2008, © Springer-Verlag 2008).