基于数字水印载体的图像语音判读方法研究

为了解决图像真伪的智能语音判读问题,提出了一种基于数字水印载体的图像语音判读方法,方法首先预存储防伪语音信息特征,而后将语音信息分解成语音符号信息以数字水印为载体嵌入到图像内部,提取复原后由专用的语音合成系统进行发音并与先前特征进行比对,进而实现判读。实验验证了方法的有效性,方法的实现使图像具备了语音校验能力和语音真伪智能判别能力。

Research on speech intelligent verification of picture based on the watermark

Aiming at the problem of the picture authentication,a method of the picture interpretation with intelligence speech based on the digital watermark is put forward.The method firstly saves the characteristic information of the target speech,and then interprets the speech to note signal and embeds it into the picture with digital watermark.After extracting and recovering,the speech can be synthesized with appropriative speech synthesis system,and then it can be compared with the original characteristics.The experiment verifies the validity of the method.