Detekcija i raspoznavanje nenormalnog ponašanja mnoštva modelom socijalnih sila

undergraduate thesis

• Title (english): Abnormal Crowd Behavior Detection and Recognition Based on Social Force Model
• Author: Marin Rađa
• Mentor: Slobodan Ribarić (mentor)
• Committee member: Slobodan Ribarić (predsjednik povjerenstva)
• Committee member: Darijan Marčetić (član povjerenstva)
• Committee member: Franjo Matković (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-09-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Model socijalnih sila prilagođen je za korištenje s česticama u video sekvencama. Računaju se interakcijske sile, iz njih se izvlače uzorci za normalne video sekvence i uzorci za nenormalne video sekvence koristeći metode vreće riječi i algoritma K srednjih vrijednosti. Video se zatim analizira koristeći metode križne korelacije da bi se pronašli normalni, tj. nenormalni, uzorci na sličici, te se na temelju glasova te sličice označavaju kao normalne ili nenormalne. Programska implementacija je testirana, te su rezultati testiranja obrađeni i prikazani. Na temelju rezultata, donesen je zaključak o primjenjivosti ovakvog modela.

Abstract (english)

Social forces model is adapted for use with particles in video sequences. We calculate interaction forces for frames, extract samples for normal video sequences and samples for abnormal video sequences using Bag of Words method and K-means algorithm. Video is then analysed using cross correlation method to find normal and abnormal patterns in the frame, and based on votes, the frame is marked as normal or abnormal. The software implementation was tested, and test results were processed and presented. Based on those test results, a conclusion was made on the applicability of such a model.

• Keywords: Računalni vid; optički tok; model socijalnih sila; napredovanje čestica; algoritam k srednjih vrijednosti; vreća riječi; vizualna riječ; latentna Dirichletova razdioba; kružna korelacija
• Keywords (english): Computer vision; optical flow; social forces model; particle advection; k means algorithm; bag of words; visual word; latent Dirichlet allocation; cross correlation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:772932
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 13:39:26