论文部分内容阅读
为了弥补空间域水印算法在提取水印时需要原宿主图像、鲁棒性较差等缺点,提出一种基于BP神经网络和人眼视觉特性(HVS)的空域数字水印算法。先选出原宿主图像对比度函数值最大的前块用于二值水印图像的嵌入,利用神经网络能够逼近任意非线性关系的特点和其自适应性,构建水印信号嵌入前后图像块像素值间的映射关系,实现了水印的盲提取,并将该算法与现有算法进行了对比。实验结果表明,对于JPEG压缩、加噪、剪切和旋转等常见的图像处理攻击,算法具有良好的鲁棒性和不可见性。
In order to make up for the shortcomings of spatial domain watermarking algorithm, such as need of original host image when extracting watermark and poor robustness, a spatial watermarking algorithm based on BP neural network and human visual characteristics (HVS) is proposed. Firstly, the first block with the largest original image contrast function value is selected for embedding binary watermarking image. The neural network can approximate the characteristics and adaptability of any non-linear relationship and construct the watermark between the pixel values before and after the embedding Mapping, the blind watermarking is achieved, and the algorithm is compared with the existing algorithms. The experimental results show that the algorithm has good robustness and invisibility to common image processing attacks such as JPEG compression, noise, cropping and rotation.