期刊名称：International Journal of Image Processing (IJIP)
电子版ISSN：1985-2304
出版年度：2015
卷号：9
期号：5
页码：272-282
出版社：Computer Science Journals
摘要：Most of the data distribution or redistribution occurs on the internet either by means of images, documents, videos, etc. But to claim the ownership and copy right protection, some extra information which cannot be removed by intruders is necessary to provide security. Such a security is provided by Watermarking. In this paper, a robust Digital Image watermarking algorithm is projected in Discrete Wavelet Transform domain using back propagation neural networks and Human Visual System Parameters like Luminous sensitivity and Texture sensitivity. Neural Networks are used in embedding and extracting the watermark. The proposed method is more protected and robust to several attacks like: Resizing, Median filtering, Row-Column copying, Low pass filtering, JPEG Compression, Rotation, Salt and Pepper Noise, Cropping, Bit Plane Removal, Blurring, Row-Column blanking, Intensity Transformation, etc. Outstanding experimental outcomes were perceived with the suggested method over a method proposed by Qiao Baoming et al. in terms of Peak Signal to Noise Ratio (PSNR) and Normalized Cross Correlation (NCC).
关键词：Image Watermarking; Discrete Wavelet Transform; Human Visual System; Back Propagation Neural Networks; Imperceptible; Robust.
