基于二次预测误差和像素排序的可逆图像水印

为了提高可逆图像水印算法的性能,提出一种二次预测误差与像素排序相结合的可逆图像水印算法;该算法首先对像素进行二次误差预测;并利用相邻像素的误差值来确定其余像素的预测误差;然后采用局部复杂度对像素进行排序,以使具有较小误差值的像素被优先处理并进行数据嵌入;最后通过平移差值直方图生成零点间隙,利用冗余的零点间隙嵌入水印信息。实验结果表明,利用二次预测误差与像素排序技术,可以有效地提高像素的预测精度并降低图像失真;在相同的嵌入容量下,该算法嵌入水印后图像的PSNR比相关算法提高了1 dB左右,具有更好的不可感知性。

Reversible image watermarking based on second-prediction-errorand pixel sorting

In order to improve the performance of the reversible image watermarking algorithm, this article proposes a reversible image watermarking algorithm based on second-prediction-error and pixel sorting. Firstly, the second-error of pixels is predicted, and the prediction error of the remaining pixels is determined by using the error of adjacent pixels. Then, the pixels are sorted by local complexity so that the pixels with the smaller error value are first processed, and the data is embedded into them. Finally, the zero gap is generated by shifting difference histogram, and the watermarking information is embedded in the zero gap of redundancy. The experimental results show that the prediction accuracy and the image distortion can be improved effectively by using the second-prediction-error and the pixel sorting technology, compared with the related algorithms, PSNR of the watermarked image is improved by about 1dB in the same embedding capacity, and the algorithm has better imperceptibility.