Digital Watermarking Using Convolutional code and Hash Function

In this paper, we propose a more robust digital watermarking method for copyright protection of digital images using convolutional code and MD5 hash function. When transmission errors caused by a noise occur, the convolutional code used in the proposed method has a good ability of detecting and correcting errors in communication channels. This method for the transmission of a signal over a noisy channel using error control coding technique considers the watermarked image as a channel and the different attacks as a noise signal in the digital watermarking. The embedded watermark in this procedure is extracted by comparing DCT coefficients of the watermarked image with DCT coefficients of the original image and then a extracted information is decoded by Viterbi algorithm and recover it to original copyright information. As a result of experiment with various scaling factor which represents a robustness of watermark, the image quality of a watermarked image is similar with existing method and it is more robust to resampling and JPEG compression like.