Ελευθερία, ΚολοκυθάEleftheria, Kolokitha2022-02-282022-02-282022-02-23http://hdl.handle.net/10889/15833In this thesis, we study the problem of recognizing/classifying different body parts from Magnetic Resonance Images (MRI). We follow an approach that utilizes the compressed sensing/sparse representations theory. Our work is based upon a recently proposed image classification method that utilizes sparse representations for the problem of face recognition. In this thesis, we propose a proper modification of that method that makes it robust against spatial translations (shifts).enClassificationMedical imagesSparse represantation theoryΤαξινόμησηΙατρικές εικόνεςΘεωρία αραιών αναπαραστάσεων