基于纠错编码的自适应图像水印算法

采用三维Lorenz映射对水印加密以提高安全性,并通过纠正水印误码来降低检测误码率,提高水印的稳健性.充分考虑了纠错编码中引入的信息冗余将导致水印嵌入强度的降低等问题,采用基于蚁群算法的图像内容边缘检测方法,使水印的分布集中在图像显著的边缘特征中,嵌入强度随着图像特征的变化而自适应地变化.在水印的检测中,利用Kalman滤波法对水印信息进行预测和估计,在没有原始图像数据的情况下,能够恢复嵌入的信息.实验结果表明,该方案具有较好的抗JPEG压缩、高斯噪声和剪切的鲁棒性.

Adaptive watermarking algorithm based on Error Correction Coding

A new encryption approach,three-dimensional chaotic Lorenz mapping,is proposed to enhance security of watermark system.To improve the robustness of watermarking,the error correcting coding is used to produce marks,which can reduce the detection error rate by correcting some errors.Considering the constraints of imperceptibility,and the decrease of the embedding strength due to the redundancy introduced by coding,a new edge detection method based on Ant Colony algorithm(ACA) is employed to constraint the embedding position to locate on salient edges,and the embedding strength also varies with the statistic of local feature coefficients of the carrier image.The watermark could be detected and estimated by Kalman filter without original data from the watermarked image,and would act as an indicator of the copyright existence.Experimental results show that the proposed scheme can improve the robustness and invisibility of watermarking and stand well against several common attacks as JPEG compression,Gauss noises and cutting.