Reversible watermarking techniques extract the watermark and recover the original image from the watermarked image without any distortion. They have been applied for those sensitive fields, such as the medicine and the military. In this thesis, a novel watermarking algorithm using multi-prediction values has been proposed. It exploits the correlation between the original pixel and the neighboring pixels to obtain twelve prediction candidates, and then selects a candidate as the prediction value according to the original pixel and the temporary prediction value. Due to the algorithm use the original pixel as one of the parameters to decide the prediction value, our prediction values are obtained with great precision. The experimental results reveal that the performance of our proposed method outperforms that proposed by Sachnev. For example the variance of the prediction errors histogram obtained by the proposed method is less than that obtained by the algorithm proposed by Sachnev et al. about 44.2%; the mean PSNR greater than about 1.47 dB and 1.1 dB under the watermark capacity 0~0.04 bpp and 0.04~0.5 bpp, respectively. Therefore, the proposed method is especially appropriate for embedding watermark in low or medium capacity. Keyword¡G reversible watermarking, watermarking, prediction, histogram shifting.

Reversible watermarking techniques extract the watermark and recover the original image losslessly from the watermarked image. They have been applied to those sensitive fields, such as the medicine and the military. Since an embedding pixel value may exceed the limitation of pixel value during the embedding process, most of the reversible watermarking methods require a location-map to record those pixels for recovering cover images. Although the location-map can be compressed by a lossless compression algorithm, and then embed into the watermarked image, this lowers embedding capacity and increases the complexity of watermarking during the procedures of embedding and extraction. In this thesis, we propose a reversible location-map free of watermarking algorithm. This algorithm first exploits the sorting and the correlation of neighboring pixels to increase the embedding capacity. Next we find thresholds from the predicted values. If the predicted value of an embedding pixel is within the thresholds, we can ensure that the pixel has no underflow or overflow problem during embedding process. Therefore, we can recover the cover image without any distortion. The experimental results reveal that the performance of our proposed method outperforms that proposed by FUJIYOSHI et al. For example, the embedding capacity obtained by the proposed method is higher than that obtained by FUJIYOSHI et al. about 60%, and the PSNR of our scheme is higher than FUJIYOSHI et al. about 5 dB.