An estimation and correction of the error of the lateral position of the vehicle within the road network based of camera data

master's thesis

• Title (croatian): Procjena i korekcija greške bočnog položaja vozila unutar cestovne mreže na temelju podataka s kamere.
• Author: Matija Mužar
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2025-02-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

The general approach to creating virtual scenarios required for validation and testing of automatic systems in vehicles typically involves equipping the vehicle with a certain sensor configuration. This sensor configuration can include cameras, radars, LiDARs etc. After the sensor configuration is picked, data is collected through real driving on certain road segments. The collected data is then processed/annotated in order to reconstruct the trajectories of moving/dynamic objects, which are then used as input data in the process of generating virtual scenarios from the recorded driving data. Throughout this process of data annotation and processing, and due to noisy sensor data trajectories can be incorrectly determined which consequently leads to possible errors in the generated scenarios. Within the scope of this Master Thesis, it is necessary to develop an algorithm for the localization of accompanying („fellow“) vehicles within the road network with an emphasis on the estimation and correction of the lateral position of the vehicle within the lane using a set of recorded camera images and GPS data. The algorithm will use AI modules for line and object detection to estimate vehicle position on the road. Then detected vehicles have to be matched with the vehicles from the recorded trajectories. Finally, a method for lateral offset estimation needs to be developed, and data has to be corrected. The algorithm needs to be verified. Aim is to reduce lateral position inconsistencies between the recorded trajectories and the corresponding road network so that in the simulation we have the same situation as it happened in the real world.

Abstract (croatian)

Generalni pristup kod kreiranja virtualnih scenarija potrebnih za validaciju i testiranje automatskih sustava u vozilima tipično se sastoji od opremanja vozila s nekom senzorskom konfiguracijom. Konfiguracije mogu uključivati, kameru, radar, LiDAR itd. Nakon što se odabere senzorska konfiguracija, podaci se prikupljaju vožnjom na stvarnim cestovnim segmentima. Podaci koji su sakupljeni su nakon toga procesuirani/anotirani kako bi se rekonstruirale trajektorije vozila u pokretu, koje se koriste kao ulazni podatak u procesu generiranja virtualnih scenarija iz snimljenih podataka. Tijekom procesa anotacije podataka i procesuiranja i zbog netočnih senzorskih podataka, trajektorije mogu biti netočno određene što dovodi do mogućih grešaka u generiranim scenarijima. Unutar ovog diplomskog rada, potrebno je razviti algoritam za lokalizaciju popratnih vozila unutar cestovne mreže s naglaskom na estimaciju i ispravljanje lateralne pozicije vozila unutar vozne trake koristeći set snimljenih slika i GPS podataka. Algoritam će koristiti module umjetne inteligencije za detekciju linija i objekata kako bi se estimirao položaj vozila na cesti. Detektirana vozila će zatim biti uparena s vozilima iz snimljenih trajektorija. Finalno, metoda za procjenu lateralnog pomaka mora biti razvijena, a podaci ispravljeni. Algoritam također mora biti verificiran. Cilj je smanjenje nedosljednosti lateralnog pomaka između snimljenih podataka i pripadajuće cestovne mreže tako da u simulaciji imamo istu situaciju koja se dogodila u stvarnome svijetu.

• Keywords: scenario-based testing; generated scenarios; annotated trajectories; lateral offset; data correction
• Keywords (croatian): testiranje s pomoću scenarija; anotirane trajektorije; lateralni pomak; ispravljanje podataka
• Language: english
• URN:NBN: urn:nbn:hr:235:907343
• Study programme: Title: Mechanical Engineering; specializations in: Design, Process and Energy Engineering, Production Engineering, Engineering Modelling and Computer Simulation, Marine Engineering, Industrial Engineering and Management, Materials Engineering, Mechanics and Robotics, Autonomous Systems, Mechatronics and Robotics; Course: Mechatronics and Robotics; Study programme type: university; Study level: graduate; Academic / professional title: sveučilišni/a magistar/magistra inženjer/inženjerka strojarstva (sveučilišni/a magistar/magistra inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2025-02-03 19:04:39