通过提出并利用相似度和平均相似度的概念,对现有的基于拐点序列码的心音识别方法的安全性进行了评估.结果表明,现有的基于拐点序列码的心音识别方法存在规律明显、易于伪造的问题.为提高心音身份识别的安全性,提出了一种改进的(S1+S2)型内部加密序列码来优化序列码结构.这种加密方法需要对一维心音信号进行小波去噪和幅值归一化处理,并对获取的原始二维心音信号进行预处理.在处理后的二维心音图中获取拐点序列码,通过将第一心音和第二心音峰值两侧对应拐点值相加形成新的序列码,降低了序列码的规律性.实验结果表明,通过平均相似度来评估二维心音图身份识别的安全性是必要的,改进的(S1+S2)型内部加密序列码能更好地提取心音的生物特征,提高了数字心音二维化识别的安全性.%The safety of existing heart sound recognition method based on the inflection point sequence code is evaluated by putting for-ward and using the concept of similar degree and average similar degree. The results indicate that there have been problems of clear rule and easy forging in the existing heart sound identification based on inflection point sequence code. In order to enhance the security of heart sound identification,a modified ( S1+S2 ) intra-encrypted sequence code is proposed to optimize the structure of sequence code. This kind of encryption method needs to conduct wavelet denoising and amplitude normalization for one-dimension of heart sound sig-nal,and preprocessing for the original two-dimensional heart sound signal which has been acquired. After that,the inflection sequence code is obtained in two-dimensional heart sound signal. The new sequence code is formed by adding value of the first heart sound and second heart sound peak corresponds to a turning point on both sides together,reduction of the rule of sequence code. The experimental results show the significance of evaluating the security and that the ( S1+S2 ) model does better in extracting biological characteristics of heart sound,and can also enhance the identification security.
展开▼
