Robusna odometrija zasnovana na fuziji GNSS-Radar-IMU senzora

master's thesis

• Title (english): Robust odometry based on fusion of GNSS-Radar-IMU sensors
• Author: Filip Vučić
• Mentor: Marija Seder (mentor)
• Committee member: Ivan Marković (član povjerenstva)
• Committee member: Juraj Peršić (član povjerenstva)
• Committee member: Luka Petrović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-11, Croatia

Abstract

Robusno i precizno poznavanje vlastitog gibanja ključan je dio autonomnih sustava. U tu svrhu uobičajeno se koriste senzori poput lasera, inercijalnih mjernih jedinica (IMU), kamera i GNSS-a. Međutim, postoje specifični scenariji u kojima se performanse ovih senzora mogu narušiti. Na primjer, laserski senzori i kamere mogu imati poteškoća u nepovoljnim vremenskim uvjetima poput magle, kiše ili dima, a GNSS signali mogu biti nedostupni. Kako bi se prevladali ti izazovi, nova istraživanja sve češće dodaju radar u postav senzora radi njegove robusnosti na vremenske uvjete. Moderni radari malih su dimenzija te energetski učinkoviti. Takvi radari generiraju oblak točaka s relativnom brzinom dobivenom Dopplerovim efektom. Stoga je moguće odrediti brzinu radara unutar pojedinog okvira točaka. IMU podaci povećavaju točnost procjene vlastitog gibanja, dok GNSS mjerenja osiguravaju dugoročnu točnost pružanjem globalnih informacija. Cilj diplomskog rada je razviti robusnu odometriju pomoću fuzije GNSS-Radar-IMU senzora te usporediti rezultate s metodom iz stanja tehnike.

Abstract (english)

Robust and precise ego-motion estimation is a key part of autonomous systems. Sensors such as lasers, inertial measurement units (IMU), cameras, and GNSS are commonly used for this purpose.    However, there are specific scenarios where the performance of these sensors can be compromised. For example, laser sensors and cameras may be challenged by adverse weather conditions such as fog, rain, or smoke, and GNSS signals can be unavailable. To overcome these challenges, novel research has explored the addition of radar to the existing sensor setup to improve weather robustness. Modern radars are compact and energy-efficient. Such radars generate point clouds with a relative speed of the points obtained by the Doppler effect. Therefore, it is possible to determine the radar ego-velocity within a single frame. IMU data increases the accuracy of ego-motion estimation, while GNSS measurements ensure long-term accuracy by providing global information. The aim of this thesis is to estimate odometry using GNSS-Radar-IMU sensor fusion and compare the results with the state-of-the-art method.

• Keywords: GNSS; Radar; Inicijalna mjerna jedinica; Fuzija senzora; Optimizacija faktorskih grafova; Odometrija
• Keywords (english): GNSS; Radar; Initial measurement unit; Sensor fusion; Factor graph optimization; Odometry
• Language: english
• URN:NBN: urn:nbn:hr:168:143749
• 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: 2023-10-11 10:58:04