基于蚁群算法的矢量量化水印

Vector quantization watermarking based on improved ant colony algorithm

提出了一种基于蚁群算法优化的矢量量化水印算法。为了用蚁群算法进行优化,首先建立了矢量量化、水印嵌入及通过噪声通道的误差模型,并给出了简化模型。在该模型中,由于水印嵌入和通道噪声直接与码字索引相关,所以该模型误差由索引值决定,又由于码字索引的确定是NP难度问题,故引入蚁群算法对其优化是方便的,优化目标是降低误码率和使不可见性达到最佳。本研究将蚁群算法引入数字水印,实验表明该优化是可行的。

A new vector quantization watermarking algorithm was proposed, At first, the error model of the vector quantization, watermarking embedding and the noisy channel were established and further the simplified model was given. Because the watermark embedding and the noisy channel were directly related to the codeword index order in the model, at the same time deciding the codeword index was a NP-hard problem typically, so the ant colony algorithm was introduced to optimize it in order to make the bit error rate lower and the embedded watermark more invisible, The ant colony algorithm is used in the watermarking field and the experiments show that the vector quantization watermarking based on the improved ant colony algorithm is successful.