Independent component analysis and its application to image processing

Independent component analysis is a new signal processing technology that shows specificity in digital image processing and so on. The principle and applications of the ICA are introduced. Mathematics model of ICA and some estimate methods are evaluated, including maximum Gaussian of ICA, minimum likelihood of ICA, and minimum mutual information of ICA. The ICA's applications in image fusion, moving object detection, face detection and feature extraction, brain signal and image analysis, digital watermark, noisy image separation are discussed. The using value and development space of ICA is fully represented.