Independent Component Analysis And Its Application Study In Multi-channel Signal Processing

Recently, theories and methods of signal processing are obtained to develope quickly. In fact, signal processing is a important foundation stone for all kinds of subjects' developing. Independent component analysis (ICA) technology developing is a later theory or method in fields of signal processing, but it was important that rapidly become a part of constitution of signal processing fields, and its developing tends gradually to maturity and systematization.This paper introduced theory and method of independent component analysis by the numbers. At the same time , it was applied to process Evoked Electroencephalography (EEG) signal which it has very important value, and to process digital image signal.This paper's contents include: brain science and analytical actuality of EEG signal, theory of ICA, research headway of evoked potential (EP) extraction technique, classes and processing algorithms of digital image watermarking, and ICA technology's application (in EEG/VEP and digital image watermarking processing), finally, summarized the work of this paper and some problems on the ICA technique which to be solved were presented and the direction of the development pointed out.The ICA technique separated successfully electromyographic (EMG) and electro-Oculographic (EOG) artifacts from EEG data ; accomplished VEP strengthening and extraction; separated watermarking and key from mixing digital image signals, even though under strong noise, the results still have very good performance index. All above showed nicer signal processing ability and factual value of ICA technique.