Πλοήγηση ανά Συγγραφέας "Vidakis, Nikolaos"

Τώρα δείχνει 1 - 2 of 2
  • Μικρογραφία εικόνας
    Τεκμήριο
    Development of a learning materials digitization platform.
    (ΕΛΜΕΠΑ, Σχολή Μηχανικών (ΣΜΗΧ), Τμήμα Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών, 2024-07-11) Iosifidis-Chokmetidis, Dimitrios; Ιωσηφίδης-Χοκμετίδης, Δημήτριος; Vidakis, Nikolaos; Βιδάκης, Νικόλαος
    The visualization of data was always one of the keyways to communicate ideas in the modern world. Most visualization models occur when big companies or even small groups of people want to view the problem in a different manner from the ordinary. Visualization is one of the most common ways to express and view those problems and solve it with the best and cost-effective solutions. Aside from metrics, graphics visualization has evolved to be one of the most fast-growing trends in terms of expressing an idea. 3d models used in different ways have proved to be invaluable to many sectors like education, manufacturing, health, and entertainment. In the last decade real time engines have improved the capabilities and introduced more dynamic experiences providing the user with a complete picture of the task ahead.
  • Μικρογραφία εικόνας
    Τεκμήριο
    Tool for automatic image segmentation and analysis of pollen grains
    (ΕΛΜΕΠΑ, Σχολή Μηχανικών (ΣΜΗΧ), ΠΜΣ Μηχανικών Πληροφορικής, 2024-04-08) Kontoulis, Vasileios; Κοντούλης, Βασίλειος; Vidakis, Nikolaos; Βιδάκης, Νικόλαος
    Beekeeping is a dynamic branch of Greek animal production, employing many beekeepers who produce significant amounts of honey. Due to various changes in the percentage of characteristics of pollen grains and limit values, it is hard to determine the botanical and geographic identity of pure honey categories, requiring continuous updates on the specifications. There is a need for new methodologies in the classification of the botanical and geographical origins of honey. Αn in-depth study of their biological actions is required to highlight their bioactive characteristics and establish a documented high nutritional value. Additionally, a collaborative platform for researchers to upload images of pollen grains can provide an accessible and efficient means for identification and analysis. This thesis proposes a web application for automatic image segmentation and analysis of pollen grains. The proposed tool allows users to upload images and execute the classification and identification of Greek pollen grains through machine-learning algorithms. To that end, this web application, serving as a tool for analysis, can communicate and wrap the functionality of third-party scripts that execute the machine learning algorithms. The application supplies the input data and gathers the results that later on are presented to the users through a REST API. The machine learning algorithms developed for this thesis using the openly accessible Cretan Pollen Dataset v1 (CPD-1) can achieve an impressive overall detection accuracy of 92%. The ultimate goal is to highlight the uniqueness of Greek honey and secure its identity, thereby enhancing its circulation both in the domestic and international markets.