基于纠错编码的回声隐藏音频水印算法

数字音频水印应用于版权保护、数字指纹、隐藏标识、保密通信等。水印误码率和归一化相关系数是衡量音频水印系统鲁棒性最重要的两个技术指标。实现方法之一的回声隐藏算法主要采用的汉明编码能有效降低水印误码率。但由于汉明码属于分组码且纠错位数只有一位,水印误码率仍然较高且对各种攻击的鲁棒性不够好。提出一种基于卷积编码的回声隐藏算法,将卷积编解码思想应用于回声隐藏算法中,利用卷积码的纠错能力降低水印误码率。结果表明:该算法较基于汉明编码的回声隐藏算法和传统回声隐藏算法降低了水印误码率,增强了音频水印系统对各种攻击的鲁棒性。

Research on echo hidden video watermarking algorithm based on error correction coding

Digital audio watermarks are used for copyright protection, digital fingerprinting, hidden identification, secure communication, and the like. Watermark error rate and normalized correlation coefficient are the two most important technical indicators to measure the robustness of audio watermarking system. The echo hiding algorithm, which is one of the implementation methods, mainly uses hamming coding to effectively reduce the watermark error rate. But since the hamming code belongs to the block code and the number of error correction bits is only one, the watermark error rate was still high and not robust enough for various attacks. This paper proposes an echo hiding algorithm based on convolutional coding, apply convolutional codec ideas to echo hiding algorithms, using the error correction capability of the convolutional code to reduce the watermark error rate. The algorithm reduces the watermark error rate compared to the echo hiding algorithm based on hamming encoding and the traditional echo hiding algorithm, enhanced the robustness of the audio watermarking system to various attacks.