English
A Fast ICA and its Application in VEP Feature Extraction
文件大小：
225k
资源说明：来自IEEE:Most biomedical signals are low frequency signals and
usually submerged by stronger noises. It is not easy to extract
their features using the normal filter methods. Independent
component analysis (ICA) in signal processing has the ability of
recovering independent source signals after they are linearly
mixed by an unknown medium. In this article we first discuss the
ICA model and the principle of extracting features of signal using
ICA. And then according the center limit theorem and the
knowledge of information theory a fast ICA algorithm based on
negentropy criterion and its implement are presented. Because of
the unavailable probability density function of the independent
source signal, its negentropy is estimated with empirical equation
which is also given in the paper. Before using ICA, it is necessary
for simplified computation to preprocess the data with removing
the mean value and whitening the data. The whole implement of
the ICA algorithm is also showed in flowchart. And lastly
applying the algorithm we successfully extract the simulated
visual evoked potential (VEP) from the mixed signals, which
shows the bright prospect of using the ICA to extract the
biomedical signal features.
本源码包内暂不包含可直接显示的源代码文件，请下载源码包。
