Repository of FERIT Osijek
Not a member yet
4053 research outputs found
Sort by
Web application for recommending and monitoring training and nutrition for athletes based on machine learning
No full textCilj završnog rada je razvoj web aplikacije za preporuku i praćenje treninga i prehrane sportaša
koji se bave trčanjem i biciklizmom. Aplikacija nudi korisnicima mogućnost procjene fizičke
spremnosti, na temelju koje se preporučuje individualizirani plan treninga i prehrane. Korisnici
mogu pratiti rezultate svojih treninga putem linijskog dijagrama, dok su konzumirani prehrambeni
proizvodi prikazani po obrocima i datumu. Osim osnovnih funkcionalnosti praćenja treninga i
prehrane, web aplikacija nudi i elemente društvene mreže. Korisnicima je omogućeno
pregledavanje profila drugih korisnika, praćenje rezultata treninga, primanje obavijesti o
treninzima i događanjima organiziranim od strane sportskih zajednica te privatna komunikacija
među korisnicima. Aplikacija je implementirana korištenjem web okvira Django, dok je korisničko
sučelje izrađeno uz pomoć HTML-a, CSS-a, Bootstrapa i JavaScript-a. Na strani poslužitelja
korišteni su SQLite baza podataka i Nutritionix API za upravljanje prehrambenim informacijama.
Razvoj modela uključuje primjenu strojnog učenja za preporuku treninga i prehrane, dok su glavni
algoritmi testirani i vrednovani kroz različite korisničke slučajeve. Dobiveni rezultati potvrdili su
ispravan rad aplikacije i njezinu upotrebljivost za krajnje korisnike.The goal of the final paper is the development of a web application for recommending and
monitoring the training and nutrition of athletes who are involved in running and cycling. The
application offers users the possibility of assessing physical fitness, based on which an
individualized training and nutrition plan is recommended. Users can monitor the results of their
training via a line diagram, while the food products they consume are displayed by portion and
date. In addition to the basic functionality of tracking training and nutrition, the web application
also offers elements of a social network. Users can view other users' profiles, track training results,
receive notifications about training and events organized by sports communities, and communicate
privately with users. The application was implemented using the Django web framework, while
the user interface was created using HTML, CSS, Bootstrap, and JavaScript. On the server side,
the SQLite database and Nutritionix API were used to manage nutritional information. The
development of the model includes the application of machine learning to recommend training and
nutrition, while the main algorithms were tested and evaluated through different user cases. The
obtained results confirmed the correct operation of the application and its usability for end users
Web application for digestive diseases risk assessment and monitoring using data analysis
No full textU ovom završnom radu osmišljena je i razvijena web aplikacija za procjenu obolijevanja i praćenja
tijeka bolesti probavnog sustava. Aplikacija je razvijena korištenjem programskog okvira Django,
programskog jezika Pythona, HTML-a i CSS-a. Ova aplikacija omogućava korisnicima
registriranje i prijavu s različitim razinama ovlasti te pristup funkcionalnostima prilagođenim
njihovoj ulozi. Pohrana osobnih podataka i praćenje simptoma oboljenja omogućeni su kroz
sigurnu bazu podataka. Analiza podatka provodi se logističkom regresijom, dok se preporuke
generiraju na temelju simptoma. Testiranjem aplikacije za različite slučajeve korištenja utvrđena
je njena točnost i učinkovitost u praćenju zdravstvenog stanja i generiranju relevantnih preporuka.In this paper, a web application was designed and developed for assessing the risk of digestive
system diseases and monitoring their progression. The application was built using the Django
framework, the Python programming language, HTML, and CSS. It allows users to register and
log in with different levels of access, providing functionalities tailored to their roles. Personal data
storage and symptom tracking are facilitated through a secure database. Data analysis is performed
using logistic regression, while recommendations are generated based on the symptoms. Testing
the application across various use cases demonstrated its accuracy and effectiveness in monitoring
health conditions and generating relevant recommendations
Business applications of smart gas meters
No full textUvođenje pametnih brojila za očitavanje plina predstavlja revoluciju u industriji energetike i pruža
niz prednosti za poslovne primjene. Kroz automatsko očitavanje podataka o potrošnji, stanju mreže
i kvaliteti energije, pametna brojila omogućuju precizno upravljanje mrežom, otkrivanje kvarova,
smanjenje gubitaka energije i proaktivno održavanje. Dodatno, pružaju personalizirane usluge
korisnicima, poput informacija o potrošnji, preporuka za uštedu energije te integraciju s pametnim
kućama. Pametna brojila omogućuju daljinsko očitavanje potrošnje, eliminirajući potrebu za
ručnim očitavanjem, što štedi vrijeme i resurse. Također, pružaju mogućnost detekcije curenja i
drugih problema u distribucijskoj mreži, što doprinosi povećanju sigurnosti i smanjenju gubitaka.
Poslovne primjene pametnih brojila za očitavanje plina obuhvaćaju optimizaciju lanca opskrbe,
automatizaciju procesa naplate, poboljšanje korisničkog iskustva i priliku za razvoj novih
poslovnih modela temeljenih na podacima o potrošnji. U konačnici, implementacija pametnih
brojila donosi mogućnost ostvarivanja značajnih ušteda i poboljšanja efikasnosti poslovanja u
industriji plina.The introduction of smart gas meters for consumption readings represents a revolution in
the energy industry and offers a range of benefits for business applications. Through automatic
consumption data reading, network status, and energy quality, smart meters enable
precise network management, fault detection, energy loss reduction, and proactive maintenance.
Additionally, they provide personalized services to users, such as consumption information,
energy-saving recommendations, and integration with smart homes. Smart meters allow remote
consumption reading, eliminating the need for manual readings, saving time and resources. They
also offer the ability to detect leaks and other issues in the distribution network, contributing to
increased safety and reduced losses. Business applications of smart gas meters include optimizing
the supply chain, automating billing processes, improving customer experience, and creating
opportunities for developing new data-driven business models based on consumption data.
Ultimately, the implementation of smart meters brings the possibility of significant cost savings
and efficiency improvements in the gas industry
Software tool for automated creation of server testing reports
No full textSpoj poslužitelja i korisničkih aplikacija postao je standard u vođenju i automatizaciji mnogih
poslovnih procesa čime se povećala važnost održavanja ispravnosti tih sustava. U ovom radu
pojašnjeni su pristupi testiranja unutar različitih metodologija razvoja programskih rješenja.
Proučava se prilagodba metodologije testiranja sve većoj potražnji za brzim i hitrim
inkrementalnim promjenama kôda sustava. Cilj rada je razviti alat s grafičkim korisničkim
sučeljem za stvaranje i pokretanje automatiziranih testnih scenarija za API poslužitelja te kreiranje
izvještaja testiranja. U svrhu razvoja vlastitog alata proučeni su neki od najpoznatijih postojećih
alata: Postman, Soap UI, Katalon studio, Apache Jmeter, Rest Assured i Karate DSL. Neke od
ključnih funkcionalnosti alata su pisanje testnih scenarija koristeći Gherkin sintaksu, automatsko
pokretanje testova, spremanje i učitavanje testova te ispis testnih izvješća. Uz pomoć pri njihovu
stvaranju, korisnici mogu vršiti kontinuirano testiranje poslužitelja i automatsku izradu testnih
izvješća. Alat se testirao nad stvarnim poslužiteljem te je uspješno izvršio dane testne scenarije i
generirao popratna izvješća.The combination of servers and user applications has become a standard in managing and
automating many business processes, which has increased the importance of maintaining the
correctness of these systems. In this paper, testing approaches within different software
development methodologies are explained. Adaptation of the testing methodology to the
increasing demand for quick and rapid incremental changes to the system code is studied. The aim
of the work is to develop a tool with a graphical user interface for creating and running automated
test scenarios for the API server and creating test reports. In order to develop our own tool, some
of the most famous existing tools were studied: Postman, Soap UI, Katalon studio, Apache Jmeter,
Rest Assured and Karate DSL. Some of the key functionalities of the tool are writing test scenarios
using Gherkin syntax, automatically running tests, saving and loading tests, and printing test
reports. With them, users can perform continuous server testing and automatically generate test
reports. The tool was tested on a real server and successfully executed the given test scenarios and
generated accompanying reports.
Ključne riječi: methodology of development of software solutions, server API testing, software
for creating test reports, Windows Forms framewor
Fingers/hand gestures detection based on non audible sound
No full textPrepoznavanje i razlikovanje gesti ruku korisnika postalo je važno zbog velikog
porasta upotrebe uređaja nosive tehnologije (engl. wearables), poput pametnog sata. Jedan od načina
detekcije gesti ruku korisnika je korištenjem nečujnog audio-signala na način da se takav signal generira
zvučnikom te se analizira reflektirani signal korištenjem mikrofona. U okviru ovog rada razvijen je
algoritam zasnovan na dubokom učenju koji koristi komercijalne zvučnike i mikrofone za generiranje
nečujnog audio-signala i analizu reflektiranog nečujnog audio-signala. Nadalje, kreirana je baza nečujnih
audio-signala gesti ruku na kojoj je neuronska mreža trenirana i testirana. Na kraju su testirane
performanse izrađenog algoritma te je implementiran za rad u stvarnom vremenu.Recognizing and differentiating user hand gestures has become important due to the large
increase in wearable technology device usage, such as smartwatches. One of the ways to detect user
hand gestures is to generate an inaudible audio signal through speakers and analize the reflected signal
using microphones. In this paper, an algorithm based on deep learning was developed, that uses
commercial speakers and microphones to generate an inaudible audio signal and analize its reflected
audio signal. Furthermore, a database of inaudible audio signals was created and used for neural
network training and testing. Finally, the algorithm's performance was tested and implemented in real
time
Mobile Android application for cooperative creation and recommendation of sightseeing tour routes
No full textU Završnom radu osmišljena je i napravljena Android aplikacija za kooperativno stvaranje i preporučivanje ruta obilaska turističkih znamenitosti koja je ostvarena u Android Studiju. Funkcionalnosti aplikacije ostvarene su programskim jezikom Kotlin, dok je opisni jezik za dizajn i izgled aplikacije korišten XML. Aplikacija imena TravelHelper korisniku daje mogućnost stvaranja profila ili registriranja, dodavanja nove znamenitosti u bazu podataka, izvlačenja podataka radi prikaza istih, ocjenjivanja znamenitosti i kreiranja rute kretanja preko Google Mapsa. Kroz ocjenjene znamenitosti korisnik ima mogućnost stvarati rutu prema ocjenama i preporukama drugih korisnika aplikacije i vlastitim interesima.. Na primjerima je utvrđeno kako je aplikacija u potpunosti ispravna i funkcionalna u svim funkcijama. Cilj je aplikacije omogućiti turistu lakše pronalaženje sadržaja u nekom gradu koji ga zanima i koji bi htio posjetiti te kako bi osobe koje žive u turističkim mjestima i od turizma mogle dodati zanimljivosti svoga mjesta te na taj način privući turiste na posjet. Drugi cilj mobilne aplikacije jest pomoći turistu u odabiru sadržaja koje želi posjetit prilikom svog boravka u nekom mjestu.In the final paper, an Android application for the cooperative creation and recommendation of tourist sightseeing routes was designed and built, which was realized in the Android Studio. The functionality of the application was realized using the programming language Kotlin while the descriptive language used for the design and appearance of the application was XML. The TravelHelper application gives the user the ability to create a profile or register, add a new landmark to the database, retrieve data for display, rate landmarks and create a route via Google Maps. Through the rated sights, the user has the possibility to create a route according to the ratings and recommendations of other users of the application and their own interests. The example show that the application is completely correct and functional in all functions. The goal of the application is to make it easier for tourists to find content in a city that they are interested in and would like to visit, and so that people who live in tourist places and from tourism can add interesting things to their place and thus attract tourists to visit. Another goal of the mobile application is to help the tourist in choosing the content he wants to visit during his stay in a place
Comparison of energy sources for the production of thermal energy in central heating systems on the example of an average residential unit
No full textOvaj rad analizira različite energente koji se koriste u centralnim toplinskim
sustavima, s posebnim naglaskom na troškove, ekološke aspekte i dugoročnu isplativost. Analizirani
energenti uključuju pelete, drva, lož ulje, prirodni plin i električnu energiju. Korištene metode uključuju
proračun troškova grijanja stambene jedinice te usporedbu emisija stakleničkih plinova za svaki
energent. Zaključeno je da biomasa i prirodni plin predstavljaju ekonomski i ekološki najisplativija
rješenja, dok električna energija, iako najskuplja, ima najmanji utjecaj na okoliš.This paper analyzes energy sources used in central heating systems, with a particular focus on
costs, environmental aspects, and long-term profitability. The analyzed energy source include pellets,
wood, heating oil, natural gas and electricity. The methods used involve calculating heating costs for a
residental unit and comparing greenhouse gas emissions for each energy source. The conclusion is that
biomass and natural gas are the most cost-effective and environmentally friendly solutions, while
electricity, although the most expensive, has the least environmental impact
Automation of the distribution network using remotely controlled devices
No full textOvaj diplosmki rad prikazuje proces automatizacije distribucijske mreže. Detaljno su
opisani glavni dijelovi distribucijske mreže kao što su daljinski upravljive rastavne naprave i prekidači te
upravljački dio. Prikazani i opisani su i softverski paketi kao što su WSOS i SCADA te ostala
komunikacijska infrastruktura. Opisana su provedena funkcionalna ispitivanja uređaja prije puštanja u
rad. Objašnjeni su razni načini automatizacije i prikazani na danim primjerima.This thesis shows the automation process of the distribution network. The main parts of the
distribution network are described in detail, such as remotely controlled disconnecting devices and
switches and the control part. Software packages such as WSOS and SCADA and other communication
infrastructure are also presented and described. Functional tests of the device before commissioning are
described. Various methods of automation are explained and shown with examples
Web application for calculating loan installments and repayment plan
No full textsažetak (hrvatski): Web-aplikacija za izračun rate kredita te otplatnog plana izrađena je koristeći web
tehnologije HTLM, CSS i Javascript uputpunjene bibliotekama Bootstrap i Font Awesome. Struktura
aplikacije definirana je HTML kodom dok je izgled prilagođen za različite veličine zaslona upotrebom CSS
koda. Javascript se koristi za izračun podataka otplatnog plana i za ažuriranje korisničkog sučelja prilikom
odabira parametra kredita. Teorijski su objašnjeni osnovni ekonomski pojmovi vezani za strukturu
kredita kao i korištene tehnologije. Praktičnim primjerom demonstrirano je korištenje web sučelja na
temelju informativnog izračuna stambenog i gotovinskog kredita. U konačnici su opisani ključni dijelovi
programskog koda i njihov utjecaj na funkcionalnosti i izgled aplikacije.The web application for calculating loan installments and repayment plans is developed using
web technologies HTML, CSS, and JavaScript, complemented by the Bootstrap and Font Awesome
libraries. The structure of the application is defined by HTML code, while the appearance is adapted for
different screen sizes using CSS code. JavaScript is used to calculate repayment plan data and update the
user interface when selecting loan parameters. The basic economic concepts related to the structure of
loans and the technologies used are theoretically explained. A practical example demonstrates the use
of the web interface based on an informative calculation of housing and cash loans. Finally, the key parts
of the program code and their impact on the functionality and appearance of the application are
described.
keywords: Loan, interest, annuity, loan calculator, web technologies, HTML, CSS, JavaScript
Web application for home repairs
No full textOvaj rad predstavlja način izrade web aplikacije za kućne servise i primjenu navedenih
tehnologija poput Angular-a, Firebase-a i JavaScript-a. Tehnologije su povezane na način što omogućava
funkcije poput klijent-server komunikacije, pretraživanje servisa, traženje servisa, kreiranje i uređivanje
servisa i prikaz servisa.This work presents a method for developing a web application for home services and the
application of technologies such as Angular, Firebase, and JavaScript. These technologies are integrated
in a way that enables functions such as client-server communication, service searching, service creation
and editing, and service displa