prikaz prve stranice dokumenta Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme
Access restricted to authenticated users
master's thesis
Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme
2021. urn:nbn:hr:179:687089
Šolić, Antonio Josip
University of Split
Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture
Department of Electronics and Computing

Cite this document

Šolić, A. J. (2021). Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme (Master's thesis). Split: University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture. Retrieved from https://urn.nsk.hr/urn:nbn:hr:179:687089

Šolić, Antonio Josip. "Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme." Master's thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, 2021. https://urn.nsk.hr/urn:nbn:hr:179:687089

Šolić, Antonio Josip. "Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme." Master's thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, 2021. https://urn.nsk.hr/urn:nbn:hr:179:687089

Šolić, A. J. (2021). 'Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme', Master's thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, accessed 17 November 2024, https://urn.nsk.hr/urn:nbn:hr:179:687089

Šolić AJ. Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme [Master's thesis]. Split: University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture; 2021 [cited 2024 November 17] Available at: https://urn.nsk.hr/urn:nbn:hr:179:687089

A. J. Šolić, "Prepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme", Master's thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, Split, 2021. Available at: https://urn.nsk.hr/urn:nbn:hr:179:687089

Please login to the repository to save this object to your list.
Metadata
TitlePrepoznavanje objekata na slikama snimljenim bespilotnom letjelicom korištenjem umjetne inteligencije i CUDA platforme
Title (english)Object recognition within images taken by a drone using artificial intelligence and the CUDA platform
AuthorAntonio Josip Šolić
MentorSven Gotovac (mentor)
GranterUniversity of Split
Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture
(Department of Electronics and Computing)
Split
Defense date and country2021-09-24, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Prepoznavanje predmeta jedna je od naglašenih primjena tehnologije dubokih neuronskih mreža i tema na koju je, posljednjih sedam godina osobito usredotočena znanstvena zajednica, kao i sve veći broj tehnoloških tvrtki i organizacija. Usporedno s tim, zadnjih godina dogodio se i uzlet tehnologije bespilotnih letjelica, kojima on otvara razne vidike primjena, što pruža nove prilike za napredak i razvoj novih tehnologija, ali i postavlja određene prepreke svojim karakterističnim zahtjevima naročito zahtjevima memorije, potrošnje energije, brzine kretanja i veličine objekata koji su snimljeni korištenjem dronova. Ovaj rad opisuje ukratko postojeće metodologije prepoznavanja predmeta, navedenu problematiku njegove primjene kod tehnologije bespilotnih letjelica uz konačan primjer primjene jednog od postojećih modela na ručno izrađenom skupu podataka korištenjem metode prijenosa znanja odnosno finog naštimavanja modela.
Abstract (english)
Object recognition is one of the most prominent applications of deep neural network technology and a topic on which the scientific community, as well as a growing number of technology companies and organizations, have been particularly focused in the last seven years. At the same time, in recent years there has been a rise in drone technology, which opens various aspects of applications, which provides new opportunities for progress and development of new technologies, but also poses certain obstacles to its characteristic requirements, especially memory, energy consumption, speed and the size of objects captured using drones. This paper briefly describes the existing object recognition methodologies, the problems of its application in drone technology with a final example of the application of one of the existing models on a handmade dataset using the method of knowledge transfer or fine tuning of the model.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:179:687089
Study programmeTitle: Computing
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsAccess restricted to authenticated users
Terms of useLicence In copyright icon
Created on2022-07-06 08:05:15