Item request has been placed! ×
Item request cannot be made. ×
loading  Processing Request

Vizijski sustav za klasifikaciju objekata temeljen na YOLOv8 modelu ; A vision system for object classification based on the YOLOv8 model

Item request has been placed! ×
Item request cannot be made. ×
loading   Processing Request
  • معلومة اضافية
    • Contributors:
      Stipančić, Tomislav
    • بيانات النشر:
      Sveučilište u Zagrebu. Fakultet strojarstva i brodogradnje.
      University of Zagreb. Faculty of Mechanical Engineering and Naval Architecture.
    • الموضوع:
      2024
    • Collection:
      Croatian Digital Theses Repository (National and University Library in Zagreb)
    • نبذة مختصرة :
      Računalni vid je područje umjetne inteligencije, odnosno dubokog učenja, koje računalnim sustavima daje mogućnost vida. Za primjenu računalnog vida su razvijeni algoritmi za izvršavanje zadataka na slikama i videima, poput prepoznavanja objekata, klasifikacije i praćenja objekata. U radu je iznesena teorijska osnova potrebna za razumijevanje algoritama računalnog vida te je potkrijepljena primjerom u kojem je korišten algoritam YOLOv8. Algoritam je treniran i provjeren na označenom skupu podataka, a zatim testiran za prepoznavanje, klasifikaciju i praćenje na neviđenom primjeru. ; Computer vision is an area of artificial intelligence, or rather deep learning, that gives computer systems the ability to see. Computer vision algorithms have been developed for performing tasks on images and videos, such as object recognition, classification and object tracking. The paper presents the theoretical basis necessary for understanding computer vision algorithms and is supported by an example in which the YOLOv8 algorithm was used. The algorithm was trained and validated on a labelled dataset and then tested for recognition, classification, and tracking on an unseen example.
    • File Description:
      application/pdf
    • Relation:
      https://zir.nsk.hr/islandora/object/fsb:10087; https://urn.nsk.hr/urn:nbn:hr:235:873965; https://repozitorij.unizg.hr/islandora/object/fsb:10087; https://repozitorij.unizg.hr/islandora/object/fsb:10087/datastream/PDF
    • الدخول الالكتروني :
      https://zir.nsk.hr/islandora/object/fsb:10087
      https://urn.nsk.hr/urn:nbn:hr:235:873965
      https://repozitorij.unizg.hr/islandora/object/fsb:10087
      https://repozitorij.unizg.hr/islandora/object/fsb:10087/datastream/PDF
    • Rights:
      http://rightsstatements.org/vocab/InC/1.0/ ; info:eu-repo/semantics/openAccess
    • الرقم المعرف:
      edsbas.73DD6BFD