Πλοήγηση ανά Συγγραφέας "Melissianos, Vasileios"
Τώρα δείχνει 1 - 1 of 1
Αποτελέσματα ανά σελίδα
Επιλογές ταξινόμησης
Τεκμήριο Categorization of medical images with modern techniques of visual and image processing methods.(ΕΛ.ΜΕ.ΠΑ., ΣΧΟΛΗ ΜΗΧΑΝΙΚΩΝ (ΣΜΗΧ), Τμήμα Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών, 2020-05-07) Melissianos, Vasileios; Μελησσιανός, ΒασίλειοςMammogram-based diagnosis of breast cancer is hard especially in high tissue density areas of the breast. In addition, Additionally, breast density type has been linked with increased risk of breast cancer calling for automated breast-density scoring tools. In this thesis, we aim to automatically classify breast density for the purpose of assisting the clinician to assess high density breasts that potentially have hidden neoplasms. The proposed methodology involves a variety of advanced techniques such as Otsu thresholding, K-Means clustering etc., for image processing, Gabor Filters and LBP for texture analysis, HOG, SURF, and SIFT for orientation/salient point extraction analysis and machine learning algorithms (for feature selection (NCA), model training (LDA) and prediction(LDA)). The results of our method using several configurations are presented for the MIAS public mammogram database.