prikaz prve stranice dokumenta NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA
No public access
undergraduate thesis
NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA
Zagreb: Algebra University College, 2016. urn:nbn:hr:225:898018
Krpan, Lovro
Algebra University College

Cite this document

Krpan, L. (2016). NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA (Undergraduate thesis). Zagreb: Algebra University College. Retrieved from https://urn.nsk.hr/urn:nbn:hr:225:898018

Krpan, Lovro. "NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA." Undergraduate thesis, Algebra University College, 2016. https://urn.nsk.hr/urn:nbn:hr:225:898018

Krpan, Lovro. "NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA." Undergraduate thesis, Algebra University College, 2016. https://urn.nsk.hr/urn:nbn:hr:225:898018

Krpan, L. (2016). 'NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA', Undergraduate thesis, Algebra University College, accessed 20 April 2024, https://urn.nsk.hr/urn:nbn:hr:225:898018

Krpan L. NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA [Undergraduate thesis]. Zagreb: Algebra University College; 2016 [cited 2024 April 20] Available at: https://urn.nsk.hr/urn:nbn:hr:225:898018

L. Krpan, "NEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA", Undergraduate thesis, Algebra University College, Zagreb, 2016. Available at: https://urn.nsk.hr/urn:nbn:hr:225:898018

Please login to the repository to save this object to your list.
Metadata
TitleNEURONSKE MREŽE I NJIHOVA PRIMJENA U RASPOZNAVANJU BILJAKA
AuthorLovro Krpan
MentorMarko Velić (mentor)
GranterAlgebra University College
Zagreb
Defense date and country2016-03-22, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
U ovom su radu opisane umjetne neuronske mreže, od jednostavnijih do nešto složenijih modela. Najznačajnije mreže u raspoznavanju slika su konvolucijske neuronske mreže te su one bile korištene u prikazivanju praktične primjene same mreže. U radu je napravljen i opisan najuspješniji istrenirani model koji je u mogućnosti napraviti kvalitetnu predikciju za raspoznavanje lista peršina i lista lovora. Veći dio skupa podataka je samostalno napravljen dok je mali dio, u svrhu validacije, preuzet sa ImageNeta. Za proces pripreme podataka i samog treniranja modela korišteni su trenutno najmoderniji alati DIGITS i Caffe, koji su također opisani u radu.
Abstract (english)
This paper describes artificial neural networks from their basic models to the more complex versions. The most important networks which are used for image recognition are convolutional neural networks, therefore they were used to demonstrate the practical application of the network itself. The paper also presents the creation and description of the most successful trained model which is capable of producing a quality prediction for distinguishing between a parsley or laurel leaf. The majority of samples were produced independently, whereas a smaller part was taken from ImageNet in order to validate the model. The tools DIGITS and Caffe, which are the most modern tools currently available and which are also described in the paper, were used in the process of data preparation and also for the testing of the model itself.
Keywords
Languagecroatian
URN:NBNurn:nbn:hr:225:898018
Study programmeTitle: Computer Engineering - Undergraduate Professional Program
Study programme type: professional
Study level: undergraduate
Academic / professional title: stručni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (stručni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
RepositoryAlgebra Univerity College
Created on2019-04-12 16:24:29

(©) Nacionalna i sveučilišna knjižnica u Zagrebu 2015. Sva prava pridržana. | Hrvatske bratske zajednice 4 p.p. 550, 10000 Zagreb, Hrvatska Tel. ++ 385 1 616-4040
OIB: 84838770814 | IBAN: HR66 2340 0091 1000 1008 1, Porezni broj (PDV identifikacijski broj) - VAT Number: HR84838770814
Prijava u repozitorij