Comparison of Several Representative Extraction Methods and Its Application to Face Recognition

Feature extraction is an important issue for face recognition.Nowdays,there are a lot of representative feature extraction methods,For example,principal component analysis and singular value decomposition,independent component analysis and non-negative matrix factorization.From point of view of algebra,the paper analyses the design principles and application charateristics of these representative methods.Singular value decomposition are methods based on matrix transformatio and independent component analysis.Non-negative matrix factorization are methods based on matrix decompose.Finally,the experimental results based on ORL and YALE face database prove its usefulness.