Search

U ovom radu riječ je o primjeni entropijskog kodiranja i wavelet transformacije na slike u boji. Cilj ovog rada je dobiti praktično primjenjivi sustav kodiranja slike koji bi imao stupanj kompresije u rangu JPG kodiranja. Na slike u boji je najprije primjenjena wavelet transformacija i nad dobivenim podacima provedene neke optimizacije odnosno aproksimacije i zaokruživanja. Tako dobiveni podaci prosljeđuju se koderu koji kodiraju te podatke i ti podaci se pohranjuju u datoteku na disk. Provedeno je i dekodiranje i rekonstrukcija slike kako bi mogli vidjeti kvalitetu rekonstruirane slike. Dokumentirane su dobivene veličine podataka u nekoliko koraka i komentirani krajnji rezultati. In this paper we talk about the application of entropy coding and wavelet transformation to colored images. The aim of this paper is to obtain a practically applicable image coding system that would have a degree of compression in the range of JPG coding. Wavelet transformation was first applied to color images, and some optimizations, approximations and roundings were performed on the obtained data. The data thus obtained is forwarded to the encoder that encodes that data, and that data is stored in a binary file on disk. Image decoding and reconstruction was also performed to be able to see the quality of the reconstructed image. The obtained data sizes in certain steps are documented and the final results are commented.

U članku se dovode u vezu tri suvremena romana kojima je nadahnuće lik Judite ili Marulićeva Judita (Pokora Pavla Pavličića, Judita Mira Gavrana i Marulov san Dina Milinovića). Autor pokušava pokazati koje su glavne kompozicijske dominante u tim knjigama, kako se prikazuju likovi te kakav se, zahvaljujući tome, postiže književni učinak, i kakav to može imati utjecaj na hrvatsku književnost i kulturu. Tematiziranje Judite i Judite pokazuje da je posrijedi tema koja je prisutna u hrvatskoj književnoj produkciji, ali ne u kanonskom čitanju, nego naprotiv u novim okvirima koji otvaraju prostor za nova kodiranja, za nove imaginacije i stilizacije. Iako se u svakom romanu postiže drukčiji učinak (u Gavranovu je posrijedi univerzalizacija lika, kod Pavličića propitivanje značenja i rekontekstualizacija hrvatske imaginarne mape, a kod Milinovića cementiranje povijesnih predodžbi), naglašava se da ih povezuje prilagodba kriminalističkom i ljubavnom žanru., The paper analyses three contemporary Croatian novels in which the figure of Judith or the Marulić's epic Judita appears: Pokora (Penance) by Pavao Pavličić (1998), Judita (Judith) by Miro Gavran (2001) and Marulov san (Marul's Dream) by Dino Milinović (2019). The objective is to demonstrate the main lines of the composition in the books, how the figures are presented and what literary effect is obtained as a result. The figure of Judith, then, and the epic Judita are present in Croatian literary fiction, but not in the canonical interpretation, rather encoded anew, novel kinds of imagination and stylisation thus being created. Pokora is a piece of crime fiction, a thriller, with love story elements, which focuses on the magical capacities of what is alleged to be the first (imagined, of course) edition of Judita: close contact between this book and other books results in these other works having their text erased from them. Various people and institutions are questing for this uncommon publication, including various institutions some of which are oriented in Croatian culture and politics to the local tradition, some to European culture. Judita by Miro Gavran, picking up directly from the Bible story and not from Marulić’s epic, radically recodes Judith. Having embarked on a love affair with Holofernes, becoming inward with her own womanliness, Judith examines her responsibility to her homeland. At the end she does in fact kill Holofernes, so carrying out her patriotically determined intention, but her hesitation modifies the character – she is turned from heroine into woman. The emphasis of the story is also modified: from collective to individual, from social to psychological. The ethics of Judith’s act, so crucial in the Bible and in Marulić is here tested out and made complex. The action of Milinović’s crime novel unfolds in Marulić’s lifetime, at the time that Judita was being composed. In Split, the Venetian authorities were doing their best to pillage the ancient heritage, and prevent the Split upper class from complaining about the Turks. The people of Split, including of course Marulić, are presented as patriots who honourably express their love of country, not only the municipal but also the national version. The figure of Marulić, for example, is predetermined by the roles that would be attributed to him by later generations. An important role in Milinović’s novel is played by Chakavian speech, rather, that amalgam of Chakavian and Shtokavian in which the urbolect of the time was stylised. But it is not used by Marulić and the Split patricians, only by the commoners. This Split idiom is marked as vernacular, as the language of the lower classes, which is strange, since Marulić’s poem is actually written in Chakavian. Analysis reveals that these three novels are differentiated by many of their features. In Gavran, the action is located in Biblical time; in Milinović in the age of Marulić himself; and in Pavličić in the present. The space of the action is, respectively, Bethulia, Split, Zagreb and the whole of Croatia. While the principal figure in Gavran is Judith, in Milinović it is the author of Judita, and in Pavličić, the poem’s readers. The three books have here different effects. In Gavran the figure of Judith is universalised by Judith being released from the role assigned her by Marulić or, if he had no such intention, the role ascribed her by later interpretations (there has been a neutralisation of the allegory). In Pavličić the emphasis is shifted from the character to the poem about the character, only to some extent to the Biblical Judith, the meanings of the book being accordingly investigated, a space being opened for the re-contextualisation of the imaginary geography of Croatia (it expands, from Split to the area of the whole of Croatia). Milinović, placing his action in the age of Marulić, reinforces the existing impression about Marulić, Judita and Split in the 16th century (about the Turkish threat, the arrogance of the Venetians, the honourableness of the Split elite). It is interesting that all these novels are adapted to popular genres, to the crime or the love story. The Marulić epic, which belongs to literary writing, is thus de-mythologized.

Ovaj rad opisuje upotrebu gem5 diskretne platforme za simulaciju računalnih sustava te u detalje objašnjava proces oblikovanja sustava za simulaciju koji koristi računalne komponente specifičnih karakteristika. Rad nastoji putem gem5 platforme simulirati kontrolni sustav što točnije modeliran po postojećem procesoru ARMv8 arhitekture naziva Cortex-A34 te zatim mijenjajući ključne karakteristike tog sustava analizira utjecaj pojedinih parametara na procese obrade video sekvenci korištenjem aplikacije za obrade video sadržaja FFmpeg. Na temelju rezultata brojnih simulacija s različitim parametrima rad iznosi zaključke o važnosti pojedinih računalnih komponenata te predlaže vrijednosti parametara sustava za što efikasniju obradu video zapisa. This paper describes the usage of the gem5 discrete platform for computer system simulation and it clarifies to an extent the process of designing simulation systems that use computer components with specific characteristics. The paper attempts to simulate a control system that emulates the real world ARMv8 architecture Cortex-A34 processor to a very close degree of similarity to then use as a reference point for many simulations with differing key characteristics in an effort to analyse the effect of certain parameters on the efficiency of video encoding and editing processes performed by the video editing software FFmpeg. Based on the results of the numerous simulations with different parameters the paper brings forth conclusions on the importance of specific computer components and suggests parameter values for the most efficient video editing procedures.

Cilj ovog završnog rada bio je analizirati neke metode moderne kriptografije. Ukratko je opisana povijest kriptografije. Dan je pregled osnovnih pojmova koji se koriste u kriptografiji te su opisane neke metode klasične kriptografije. Dana je podjela modernih kriptosustava te su neki najpoznatiji opisani u radu. U radu je opisano na koji način se u praksi koriste kriptografske metode i kako pomoću njih zadovoljiti sigurnosne zahtjeve komunikacijskih sustava. U praktičnom dijelu rada implementirani su DES, RSA i Diffie-Hellmanov protokol za razmjenu tajnog ključa u programskom jeziku Python. The goal of this bachelor thesis is to analyze some of the algorithms of modern cryptography. The history of cryptography is briefly described, a review of basic concepts of cryptography is given, and some methods of classical cryptography are described. Also, a classification of cryptosystems is given and some of the more known cryptosystems are described. It's explained how modern cryptographic algorithms are applied in the real world and how to satisfy the security requirements of communication systems using these algorithms. The practical part of this bachelor thesis gives an implementation of DES, RSA, and the Diffie–Hellman key exchange protocol in the programming language Python.

U ovom radu razmatran je utjecaj procesora, cache memorije i vrste RAM-a na performanse kodiranja videa. To je izvedeno simulacijama koristeći gem5-X simulator na superračunalu Bura u Rijeci. Simulacijama su dobiveni rezultati koji su sukladni teoriji koja stoji iza svega toga. Pokazano je kako povećanje veličine cache memorije pozitivno utječe na performanse kao i povećanje asocijativnosti cache memorije. Prilikom razmatranja utjecaja procesora utvrđeno je kako povećanje frekvencije takta poboljšava performanse. Isto tako, tip RAM-a koji ima veću frekvenciju daje bolje rezultate zbog veće propusnosti. Svaka od ovih komponenti ima gornju granicu do koje možemo povećavati razmatrane parametre zbog cijene ili kompleksnosti izvedbe. Kod simulacija s različitim brojem dretvi, rezultati su pokazali kako program s više dretvi ima bolje performanse. In this paper was discussed the impact of processor, cache and type of RAM on performance of encoding video. To make simulations, the gem5-X simulator was used on Bura supercomputer in Rijeka. The simulations provided the results that are consistent with the theory behind all of this. It has been shown that increasing the size of the cache memory has a positive effect on performanse as well as increasing the associativity of the cache memory. When considering the impact of the processor, the results showed that increasing the clock frequency improves performance. Likewise, the type of RAM that has a higher frequency gives better results due to higher bandwidth. Each of these components has an upper limit to which we can increase the considered parameters due to cost or complexity of performance. In simulations with different number of threads, the results showed that a program with more threads has better performance.

Ovaj završni rad opisuje matematičke osnove kriptografije i njene povijesti. Pošiljatelj šifrira poruku i takvu šalje primatelju koji istu dešifrira na svom uređaju i čita izvornu poruku. Enkripcijski algoritmi mogu se podijeliti na simetrične i asimetrične. Rad posebno analizira, implementira i koristi RSA i AES enkripcijske algoritme na primjeru mobilne aplikacije za zaštićenu razmjenu poruka za operacijski sustav Android. Prilikom implementacije navedenih algoritama, svaka tehnička komplikacija detaljno je analizirana i objašnjena. Aplikacija omogućava stvaranje korisničkog računa te slanje tekstualnih i slikovnih poruka drugim korisnicima aplikacije. Svaka poruka poslana preko izgrađene aplikacije šifrirana je navedenim enkripcijskim algoritmima cijelim komunikacijskim kanalom. This undergraduate dissertation describes the mathematical basis and history of cryptography. The sender encodes the message and sends it to the recipient, who decodes the message on their device and reads the original message. Cryptosystems are divided into the symmetric and the asymmetric. The paper particularly analyses, implements and uses RSA and AES cryptosystems in the example of a mobile application for secure messaging for Android operating system. During the implementation of the mentioned algorithms, each technical complication was analysed and explained in detail. The application allows you to create an account and send text and image messages to other users. Each message sent via the built application is encoded with the mentioned cryptosystem throughout the communication channel.

U ovom završnom radu RFID TEHNOLOGIJA opisano je funkcioniranje te primjene i mogućnosti RFID tehnologije. RFID ili Radiofrekventna identifikacija je tehnologija brzog razvoja koja je zastupljena u svim područjima. Koristi se za stotine, ako ne i tisuće primjena današnjice. U prvom dijelu rada opisan je RFID sustav, rečeno je kako se svaki RFID sustav sastoji od transpondera, čitača i antene te je tu opisan princip rada i karakteristike svakog od dijelova sustava. U trećem poglavlju opisane su frekvencije rada sustava, LF, HF, UHF i Mikrovalovi. U nastavku je opisan izbor frekvencije rada, modulacija te kodiranje signala, poput NRZ, Manchester i Diferencijalnog kodiranja. U posljednjem dijelu opisana je primjena ove tehnologije te postupci uspješne zaštite., In this final assignment, the RFID TECHNOLOGY describes the operation and application and capabilities of the RFID technology. RFID or Rad io Frequency Identification is a fast -paced technology that is prevalent in all ares. It is used for hundreds if not hhousands of applications today. The first part of the final assignment describes the RFID system, says that each RFID system consists of a transponder, a reader and an antenna, and describes the principle of operation and characteristics of each of the parts of the system. In the third chapter, the operating frequencies of the essence, LF, HF, UHF and Microwave frequency are described. The following describes the frequency of choice, modulation and signal encoding, such as NRZ, Manchester and Differential coding. The last section describes the application of this technology and the procedures for successful protection.

Dubinske kamere postaju sve dostupnije i korištenije. Problem dubinskih kamera je generiranje velike količine podataka. U ovom radu predstavljena je implementacija rješenja problema korištenjem trenutno dostupnog standarda za kompresiju H.264. Dubinska informacija se kodira kako bi se prilagodila video kompresiji te se potom s RGB video slikom kompresira H.264 standardom. Time se postiže značajna ušteda memorijskog prostora i omogućava jednostavniji prijenos trodimenzionalne informacije postojećom mrežnom infrastrukturom. Provedeno je iskustveno ispitivanje na malom broju korisnika. Rezultati iskustvenog ispitivanja su pokazali da nema značajnijeg gubitka u kvaliteti video zapisa umjerenim korištenjem kvantizacijskog parametra. Depth cameras have become commonly available and used. The problem depth camera has is generation of very large amounts of data. This thesis introduces the implementation of the solution for the problem by using already available compression standard H.264. Depth information is first coded so it would be better prepared for the video compression and then it is, together with the RGB video data encoded by the H.264 standard. By doing so, required memory space is significantly reduced and easier network transfer of the three-dimensional information over the current infrastructure is accomplished. The testing of user experience is conducted on a small number of people. Results of the testing showed that there is no significant quality loss when using moderately high value of quantization parameter.

U radu je prikazan način rada video kompresije. Postoje dvije vrste video kompresije: lossless i lossy. Detaljno je opisan postupak kodiranja video sadržaja koju koristi JPEG norma. Kodirati se može na dva načina: kodiranje unutar slike iz kojeg proizlaze I-slike i kodiranje između slika koje rezultira s P-slikama i/ili B-slikama. Također, opisane su razne norme za kodiranje videa. Neke od njih su: H.261, H.263, MPEG-2, H.264/AVC (MPEG-4). Najpoznatije i u širokoj upotrjebi su H.264/AVC i MPEG-2. Najveća prednost H.264/AVC nad MPEG-2 je postizanje iste kvalitete video sadržaja s od 2 do 2,5 puta manjom brzinom prijenosa (točnost određivanja vektora pomaka je puno veća). Nadalje, uspoređivanjem svih normi po istom kriteriju, ispostavilo se da je H.262 norma s najvišom brzinom prijenosa. Osim toga, opisani su razni algoritmi za procjenu pokreta. Na samom kraju opisana je i analizirana metoda kojom su se tražile sličnosti između dva videa pomoću vektora pomaka. This paper describes the way video compression works. There are two types of video compression: lossless and lossy. The process of encoding video content using the JPEG standard is described in detail. Video content can be encoded in two ways: intra-frame coding which produces I-images and inter-frame coding which produces P-images and/or B-images. Also, various video coding standards are described. Some of them are: H.261, H.263, MPEG-2, H.264/AVC (MPEG-4). The most popular and widely used are H.264/AVC and MPEG-2. The biggest advantage of H.264/AVC over MPEG-2 is achievement of the same video quality with 2 to 2.5 times lower bit-rate (the accuracy of motion estimation is higher). Furthermore, by comparing all the norms by the same criterion, it turned out that the standard with the highest bit-rate is H.262. In addition, various motion estimation algorithms are described. At the very end, an analysis and description is shown of a method which uses motion vectors to find similarities between two videos.

Predmet ovog rada je neverbalna komunikacija te njen utjecaj prilikom javnog nastupa. U prvom dijelu rada govori se općenito o neverbalnoj komunikaciji odnosno komunikaciji bez riječi, bilo da je ona namjerna ili nenamjerna. Ona je uvijek prisutna kada se ljudi susreću, pojmovima koji su vezan uz ovaj oblik komunikacije točnije enkodiranju i dekodiranju. Neverbalna komunikacija svojim raznolikim funkcijama i kanalima primjene uvelike utječe na komunikaciju, razumijevanje, stvaranje dojma i uvjerljivost i povjerenje govornika. Dalje rad se bavi odnosom prezentacije i javnosti, odnosima s javnošću njegovim vrstama kako bi se mogla shvatiti šira slika kome se sve obraća govornik prilikom svog izlaganja, zatim prezentiranje u javnosti odnosno javni nastup te govor pred publikom. Object of this work is nonverbal communication and its influence during public speaking. First part of work is generaly about nonverbal communication or communication without words, deliberatey or not. It is always present when people meet, by the concepts related with this form of communication, coding and encoding. Nonverbal communication with functions and channals of use greatly influence on communication generaly, understanding, making impression, stringency and trust. This work is also about presentation in public, public relations an types of public relations so it would be easier to understand all kinds of audience , then speaking in public with some kind of purpose like informing, influencing and getting to know about some work or product.

U posljednjih nekoliko desetljeća sve se više povijesnih tekstova digitalizira. Međutim kako se razvoj odvijao izolirano na projektima, sveučilištima i institutima, ili među pojedincima, nedostajala je komunikacija među članovima zajednice, što je rezultiralo resursima, alatima i sustavima koji međusobno ne mogu razmjenjivati informacije. Stoga se u novije vrijeme pokazala potreba za standardizacijom tehnologija, ali i cijelih procesa razvoja. Kao ključni pojam u fazi digitalizacijskih projekata javlja se interoperabilnost, kojom se želi omogućiti međusobna komunikacija digitaliziranih resursa. U radu je prikazan način kodiranja rječničkih natuknica iz della Bellina rječnika iz XVIII. st. korištenjem TEI (Text Encoding Initiative) standarda te njihovo povezivanje na vanjske resurse (mrežno izdanje Hrvatske enciklopedije Leksikografskog zavoda Miroslav Krleža i hrvatsku inačicu Wikipedije)., The digitization process of historical texts is increasing in the last several decades. However, the digitization projects are usually isolated to the project teams, universities and institutes, but also to individuals developing the project. There was a lack of communication between the community members, which resulted in resources, tools and systems that are not able to exchange information. Consequently, recently it was demonstrated that there is demand for standardization of technologies, but also all processes of development. Interoperability emerges as the key concept at this phase of digital humanities, which aims to facilitate communication of data. In this paper we present encoding scheme of dictionary entries of della Bella’s 18th century dictionary using the TEI (Text Encoding Initiative) standard and a method of linking to external resources (web edition of Croatian encyclopedia of the Miroslav Krleža Institute of Lexicography and the Croatian version of Wikipedia).

Digitalni vodeni žig je vrsta markera koji je ugrađen u sliku, video i zvučni signal, tekstualni dokument, kompjutorski program, kemijski podatak... Obično se koristi u svrhu identifikacije vlasništva nad autorskim pravima sadržaja. Potreba za označavanjem sadržaja digitalnim vodenim žigom javlja se prilikom obrade podataka, preciznije njihovim prelaskom iz analogne domene u digitalnu čime postaje nemoguće razlikovati originalne podatke od kopija. Umetanje vodenog žiga radi se na način da se pomoću određenog algoritma kodiranja žig umeće u određeni dokument, te se kasnije određenim algoritmom dekodiranja žig odstranjuje iz dokumenta i tako jednoznačno utvrđuje vlasnika i porijeklo dokumenta. Digitalne vodene žigove prema vizualnoj percepciji svrstavamo u četiri skupine. Prvoj skupini pripadaju vidljivi vodeni žigovi kod kojih je karakteristično da je na izvornom dokumentu žig vidljiv kao logo oznaka. Drugoj skupini pripadaju robusni nevidljivi vodeni žigovi kod kojih je karakteristično da su vizualno nevidljivi, ali ih se može detektirati dekoderom, a ujedno su otporni i na napade. Trećoj skupini pripadaju lomljivi nevidljivi vodeni žigovi kojima je također karakteristično da je žig nevidljiv, ali ga je moguće detektirati i ranjiv je na sve vrste napada. Četvrtoj skupini pripadaju dvostruki vodeni žigovi kojima je karakteristično da su kombinacija nevidljivog i vidljivog vodenog žiga. Digitalni vodeni žigovi svoju primjenu imaju u dokazivanju autentičnosti podataka, praćenju emitiranja programa na televiziji, u situacijama u kojima dodatna informacija o digitalnom sadržaju ne treba sadržavati informacije o vlasniku nego o krajnjem korisniku, i u mnogim drugim sličnim situacijama u životu. Digital watermark is a type of marker that is embedded in an image, video and audio signal, a text document, a computer program, chemical data… Usually it is used for identification of copyright ownership. The need for watermarking occurs in data processing, more specifically during their transition from analogue to digital domain, which is when it becomes impossible to distinguish the original data from the copies. The procedure is such that the watermark is first inserted by using a particular coding algorithm. After this, a specific decoding algorithm is used to remove the watermark, to determine the unique owner and origin of the content. Digital watermarks are classified into four groups according to visual perception. The first group includes visible watermarks that are displayed as logo marks on the original document. The second group includes robust invisible watermarks that cannot be seen, but can be detected by a decoder. Another characteristic is that they are resistant to attacks. The third group are fragile invisible watermarks which are also invisible, but possible to detect and vulnerable to all kinds of attack. The fourth group are dual watermarks, which are a combination of visible and invisible watermarks. Digital watermarks have their purpose in data authentication, broadcast monitoring, in situations where additional information on digital content should not contain ownership information, but it should contain information about the end user, and in many other similar situations in life.