Robust digital image watermarking using Pseudo Random Numbers

Communication of digital data over Internet has increased drastically as huge amount of multimedia data is embedded over Internet every day. Digital images contribute a major part of multimedia data. The rapid advances in computer network hardware and software has led to high risk of piracy. Therefore, there is a need of techniques for copyright protection and content authentication of digital images. Watermarking is one such technique. This paper presents a robust image watermarking technique in wavelet domain that uses the Pseudo Random Number (PRN) sequence generated using Elman neural network to watermark the image. The proposed method decomposes original image in the wavelet domain using discrete wavelet transform and watermarks it with a PRN sequence. The watermark is extracted to recover the original image. The neural network generated PRN sequence is a highly random sequence providing high level of security. This approach shows an excellent resistance against intensity transformation attack.