Open Access
Subscription Access
Face Recognition Using Principal Component Analysis With Statistical Median and Mode for Normalization
Face Recognition is a challenging task now a days and till date, there is no technique which provides robust solutions to all situations. This Paper focus on Face Recognition by using Principal Component Analysis (PCA) with Median and Mode for Normalization. The original PCA calculates the Mean (face) data and then subtracts the Mean from the each (face) data item with an intention to Normalize the data. This concept of Normalization in PCA is repeated for Statistical Median and Mode also, the other measures of Central tendency. The results obtained are compared for three different Normalization modes of PCA i.e. Mean PCA, Median PCA and Mode PCA. Results show that Median PCA is preferred over Original Mean PCA or Mode PCA in some situations.
Keywords
Face Recognition, Principal Component Analysis (PCA), Mean PCA, Median PCA, Mode PCA, Artificial Neural Network (ANN), Support Vector Machine, Hidden Markov Model (HMM), Euclidean Distance, City Block Distance, Cosine Similarity Distance, MAHCOS Distance, Normalization.
User
Font Size
Information
Abstract Views: 165
PDF Views: 0