Web aplikacija za nadgledanje prostora temeljena na umjetnoj inteligenciji

master's thesis

• Title (english): Web application for spatial monitoring based on artificial intelligence
• Author: Nikolai Ivan Horvat
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2022-07-18, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing; Artificial Intelligence

Abstract

Računalni vid interdisciplinarno je područje koje obuhvaća mnoge grane poput inženjerstva, računarstva, matematike, ali i biologije, fizike i psihologije. Ono omogućuje računalima da vide, tj. da procesiraju i prepoznaju objekte na slici kao što to mogu ljudi. S obzirom da se dnevno objavljuje i više od četiri milijarde slika, javlja se potreba za razvojem novih i bržih algoritama koji omogućuju napredak računalnom vidu. Uzimajući u obzir današnje trendove razvoja i dizajna, u ovom je radu razvijena web aplikacija za nadzor prostora. Ona predstavlja sučelje između korisnika i sustava u kojem je prikazan pogled kamere u stvarnom vremenu i preko kojeg korisnik može pregledati videozapise u arhivi. Kao okidač spremanja videozapisa na raspolaganju su dva poznata modela strojnog učenja koja se mogu koristiti za praćenje – detekcija lica temeljena na Haar značajkama (algoritam Viola-Jones) i detekcija pokreta optical flow. Na kraju rada provedena je evaluacija aplikacije koja uključuje ljudske subjekte.

Abstract (english)

Computer vision is an interdisciplinary field that includes many areas such as engineering, computing, mathematics, but also biology, physics and psychology. It gives computer the ability to see – to process and recognize objects in an image as humans can. Given that more than four billion images are published daily, there is a need for the development of new and faster algorithms that enable computer vision to progress. Considering today's development and design trends, a web application for space monitoring was developed in this paper. It represents an interface between the user and the system in which the camera view is displayed in real time and through which the user can browse the videos in the archive. The trigger to save the videos are two of the famous machine learning models that can be used for tracking - face detection based on Haar features (Viola-Jones algorithm) and optical flow motion detection. At the end of the paper, an evaluation of the web application involving human subjects was carried out.

• Keywords: računalni vid; strojno učenje; umjetna inteligencija; web aplikacija; JavaScript; Python; OpenCV; Viola-Jones; optical flow
• Keywords (english): computer vision; machine learning; artificial intelligence; web application; JavaScript; Python; OpenCV; Viola-Jones; optical flow
• Language: croatian
• URN:NBN: urn:nbn:hr:235:017002
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka strojarstva (magistar/magistra inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2022-07-15 10:49:47