Temel Bileşenler Analizi ve Kanonik Korelasyon Analizi ile İmge Tanıma ve Sınıflandırma

Abstract

This work investigates the role of canonical correlations analysis in image recognition and classification problems with comparison to principal components analysis. Principal components analysis is a well-known and widely used feature selection and reduction method for face recognition problems. In this thesis, canonical correlation analysis is proposed as an alternative feature selection and reduction method for generic image recognition and classification problems. This new method is studied via various image recognition and classification problems in comparison with principal components. Multiple canonical correlation analysis is proposed as a new feature selection and dimension reduction algorithm for image classification problems involving multiple classes. By studying various image recognition and classification problems, it is shown that canonical correlation analysis is a more efficient methodology for feature reduction in comparison to principal components analysis.