应用时变小波包零树编码实现心电数据的高效压缩

为减少传统小波零树编码算法中孤立零的出现,以确保编码压缩后信号的熵减,提出了时变小波包零树编码算法,构建了最优小波包基下低冗余映射关系,同时考虑到心电信号自身的时域相关性,将时频域结合考虑,进一步降低了信息熵,最终明显改善了压缩效果.实验结果表明,该方法保留了小波零树编码可以实现累进传输编码的优点,并且取得了更好的压缩性能.

High Efficiency Compression of Electrocardiogram Data Based on Time-Varying Zerotree Wavelet Packet Coding

During the process of tele electrocardiogram monitoring,efficient data compression has become an important link of data processing under the demands of mass data storing and transmitting.To achieve this aim,a data compression method is proposed based on time-varying zerotree wavelet packet coding,which can improve the isolational zeros of traditional zerotree wavelet coding and realize the further decrease of the information entropy.Correspondingly,a low redundancy mapping is constructed under optimum basis of wavelet packet,and the self-similarity in time domain is also considered to decrease the information entropy,which gives the obvious improvements to the final compression results.The experiments show that this new method reserves the merit of zerotree wavelet coding on progression-transmission, and can obtain the better results in data compression.