Independent Component Analysis And Its Applications Of Image Feature Extraction And Denoising

Independent Component Analysis (ICA) is a novel signal processing method developed recently which analyzes the data from a statistical point of view. ICA aims at recovering the latent variables (sources) only from the observed mixtures. It is a generalization of Principal Component Analysis (PCA) that separates the higher-order dependencies in the observed mixtures, in addition to the second-order dependencies. Compared with the other methods, the independent components of ICA are both nongaussian and statistically independent. It is well known that, in the field of image analysis and pattern recognition, much of the important information of the images may be contained in the higher-order relationships among the image pixels. ICA. based on higher-order statistics, has shown great promising ability in image feature extraction and image compression.This thesis presents the theories and algorithms of ICA and explores its applications in the image feature extraction, image denoising and face recognition. The main work of the author focuses on the following aspects:1) First, give an overview of the status of ICA research.2) Introduce some ICA-related knowledge: statistics and information theory.3) Describe the concepts of two kinds of multidimensional data representations: PCA and ICA. and reveal their difference.4) Based on the nongaussian maximum principle and information maximum principle, we study the two typical algorithms of ICA. and write program to implement these algorithms.5) In the experiments we combine ICA with sparse coding to find an efficient coding of natural scenes. In addition we use a sliding subwindow and a thresholding operator on the coefficients to reduce the Gaussian noise. In the task of face recognition, we preprocess the data with PCA in order to reduce the number of ICs and the complexity of the problem. We also compare the ICA method with some traditional methods: PCA and Wiener filter algorithm. The comparisons show ICA has better performance in image feature extraction and denoising.