Research On Independent Component Analysis Algorithm And Its Application In Image Processing

Independent component analysis algorithm (ICA) is a recenly developed an effective blind signal processing technique, and it is playing an increasingly impornent role in many areas. ICA has important theory and application value, and it has a wide application in the speech signal、image processing、aireless conmmunations and many other areas. In the past ten years, its theory had been developed quickly and a lot of effective algorithms had been made. Currently, ICA algorithm is a hotspot in international signal processing areas.Fristly, this paper introduces the basic theory of ICA, the typical ICA algorithms and its performance analysis. The basic theories mainly include information theory, the mathematical model of ICA and its solvability analysis etc. Typical ICA algorithms mainly include maximum entropy algorithm, H-J algorithm, minimum mutual information algorithm, stochastic gradient algorithm and the natural gradient algorithm etc.Secondly, this paper further analysises the FastICA algorithm which is widely used, amends the original Newton iterative algorithm, and prorpses the "Newton Type" iteration algorithm with third-order convergence and Newton iteration algorithm with fifth-order convergence. The simulation result of the image signal separation shows that the improved algorithm is better than the conventional FastICA algorithm in the separation, can significantly reduce the times of iterations and the running time, and can also increase the convergence rate and operation efficiency.Finally, the improved FastICA algorithm is applied to image denoising, and we combination the ICA algorithms with the fliter algorithms, which make the image denosing efficiency achieve better. Simulation results show that the improved algorithm make the SNR of the denoised image increase greatly. In simulation experiment of the face recognition, the improved FastICA algorithm is used for facial expression feature extraction. The results show that the improved FastICA can effactively extract facial expression feature.