摘要
研究了心电信号的非平稳过程特性 ,从时间序列建模角度分析动态心电数据表示模型和压缩算法。采用小波网络作为建模工具 ,将原始心电数据映射为小波网络的网络参数作为数据重构信息 ,给出压缩 /重建的实验结果并分析讨论。研究表明 ,小波网络压缩算法继承了小波变换和神经网络的优点 ,具有较好的压缩性能。
The non-stationary process character of ECG is studied. The data representation model and compression algorithm is researched with time-series modeling theory.Using wavelet networks as modeling tool, a compression algorithm of ECG is proposed, which maps the original data of ECG to parameters of wavelet networks as recovery information. Then the experiment results of data compression/ recovery are referred and analyzed.From the study, the conclusion can be obtained that the compression algorithm can provide superior performance for the virtue of wavelet networks inherited from wavelet and neural networks.
出处
《西北大学学报（自然科学版）》
CAS
CSCD
北大核心
2001年第6期469-472,共4页
Journal of Northwest University（Natural Science Edition）
基金
国家自然科学基金资助项目 (199710 6 5 )
