li8504的专栏

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.

matlab Wiener滤波进行图像恢复,首先对棋盘图像进行模糊，增加加性噪声，然后利用Wiener滤波进行图像恢复。