Procjena i korekcija lateralnog gibanja vozila

master's thesis

• Title (english): Vehicles' lateral motion estimation and correction
• Author: Marko Turina
• Mentor: Marija Seder (mentor)
• Committee member: Ivan Petrović (član povjerenstva)
• Committee member: Ivan Marković (član povjerenstva)
• Committee member: Luka Petrović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-09-17, Croatia

Abstract

Ovaj diplomski rad bavi se procijenom i korekcijom lateralnog gibanja vozila u svrhu
točnijih podataka. Precizna detekcija položaja unutar prometnih traka nužna je za procjene položaja i gibanja svih sudionika u prometu na određenom segmentu ceste. Pomoću kamera, prvenstveno prednje, dobiju se videozapisi vožnje. Zadatak je, analizom
dobivenih slika putem algoritma detekcije linija prometnih traka i algoritmom za detekciju objekata odrediti i maksimalno točno ispraviti trajektorije na način da budu što
sličniji stvarnim podacima. Opisani su načini i tehnologije koje omogućavaju u stvarnom
vremenu transformirati sliku iz kamere u pogled odozgo te time otkloniti efekte izobličenja perspektive. Time je omogućeno precizno otkrivanje objekata i obilježja ceste. Rezultati pokazuju da sustav može učinkovito procijeniti i korigirati lateralni položaj vozila
u scenariju te se time približiti cilju točnosti podataka.

Abstract (english)

This master thesis deals with the estimation and correction of the lateral offset of the
vehicle for the purpose of more accurate data. Precise detection of the position within
traffic lanes is necessary to assess the position and movement of all road users on a certain road segment. Driving videos are obtained using the cameras, primarily the front
ones. The task is to determine and correct the trajectories as closely as possible to the
real data by analyzing the obtained images using the lane line detection algorithm and
the object detection algorithm. Methods and technologies are described that make it
possible to transform the image from the camera into a view from above in real time,
thereby removing the effects of perspective distortion. This enables precise detection of
objects and road features. The results show that the system can effectively estimate and
correct the lateral position of the vehicles in scenario, thereby approaching the goal of
data accuracy.

• Keywords: YOLO; YOLinO; detekcija objekata; detekcija linija traka; CNN; Python
• Keywords (english): YOLO; YOLinO; object detection; lane line detection; CNN; Python
• Language: croatian
• URN:NBN: urn:nbn:hr:168:750399
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 22:22:51