Praćenje objekata metodom globalnih najbližih susjeda

undergraduate thesis

• Title (english): Object Tracking via Global Nearest Neighbor Method
• Author: Dorian Možnik
• Mentor: Nikola Mišković (mentor)
• Committee member: Nikola Mišković (član povjerenstva)
• Committee member: Ivan Marković (član povjerenstva)
• Committee member: Đula Nađ (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-02-08, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Najveći izazov kod praćenja više objekata je što nije moguće sa sigurnošću odrediti koje detekcije pripadaju kojem objektu, a također je moguće i da nisu svi objekti detektirani u svakom trenutku. Neki od najvažnijih algoritama se bave tim problema. Među prvim koji su se pojavili je i metoda globalnog najbližeg susjeda. U ovom radu prošli smo kroz teorijsku pozadinu Kalmanovog filtra i metode globalnog najbližeg susjeda. Zatim smo kombinirali te dvije metode da bi implementirali jednostavan algoritam za praćenje. Također, na kraju, je napravljeno i proširenje pokušajem implementacije rođenja i smrti objekata metodom baziranoj na životnom vijeku detekcija.

Abstract (english)

The biggest challenge when tracking multiple objects is that it is not possible to determine with certainty which detections belong to which object, and it is also possible that not all objects are detected at all times. Some of the most important algorithms deal with this problem. Among the first of them to appear is the global nearest neighbor method. In this paper, we went through the theoretical background of the Kalman filter and the global nearest neighbor method. We then combined the two methods to implement a simple tracking algorithm. Also, at the end, an extension was made by trying to implement the birth and death of objects using a method based on the lifetime of detections.

• Keywords: praćenje objekata; globalni najbliži susjedi; Kalmanov filtar; mađarska metoda; matrica troškova
• Keywords (english): object tracking; global nearest neighbor; Kalman filter; Hungarian algorithm; cost matrix
• Language: croatian
• URN:NBN: urn:nbn:hr:168:770530
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-07-13 10:23:22