ISBN-13: 9783844302783 / Angielski / Miękka / 2011 / 188 str.
Face recognition is a difficult problem .Thedimension reduction and the skill to acquire minimumsize of feature vector required for face recognitionfor diverse facial expressions is a challenging taskin face recognition. Dimension reduction results inremoval of irrelevant variables along with noisetherein and a lower computation complexity ofsubsequent processing. This book addresses thechallenges of dimension reduction, choice of minimumsize feature vector for face recognition andminimization of adverse effects of varying facialexpressions on the recognition through reduction inimage resolution. Face images with varying imageresolution are obtained through image decimation.Improved recognition rate results are also reportedusing novel coefficients selection and optimizationmethods in DFT, DCT and DWT based face recognitionmethods. The experiments are carried out for variousimage resolutions using five different datasets.