Computer Vision for Road User Detection and Localisation in Intelligent Transportation System Infrastructure

Risto Ojala

Computer Vision for Road User Detection and Localisation in Intelligent Transportation System Infrastructure

Julkaisun otsikon käännös: Tietokonenäön menetelmiä tienkäyttäjien tunnistukseen ja paikannukseen älykkäiden liikennejärjestelmien infrastruktuurissa

Risto Ojala

Tutkimustuotos: Doctoral Thesis › Collection of Articles

Abstrakti

Älykkäille liikennejärjestelmille on kasvava tarve maailmalla. Niillä pyritään ratkaisemaan erinäisiä tieliikenteen ongelmia, kuten ruuhkia, saasteita, sekä liikenteen turvallisuushaasteita. Älykkäiden järjestelmien toiminta tyypillisesti edellyttää kattavaa ja ajantasaista tietoa liikenteestä tehokasta päätöksentekoa varten. Tässä väitöskirjassa tutkittiin teknisiä ratkaisuja, jotka mahdollistavat liikennekameroiden hyödyntämistä tienkäyttäjien havainnointiin ja paikantamiseen. Tutkimus keskittyi erityisesti ongelmiin, jotka tällä hetkellä rajoittavat tietokonenäköteknologioiden soveltamista liikennekameroissa kyseiseen tehtävään. Kehitystyö pohjautui vahvasti kokeelliseen tutkimukseen, tuottaen konkreettisia tuloksia tienkäyttäjien tunnistuksesta ja paikannuksesta liikennekameroilla. Osana tutkimusta kehitettiin suorituskyvyltään äärimmäisen nopea tienkäyttäjien tunnistusmenetelmä liikennekameroille. Menetelmä perustuu taustan tunnistukseen ja konvoluutioneuroverkkoon pohjautuvaan luokittelijaan. Käytännön kokeissa menetelmä kykeni käsittelemään yli 30 kuvaa sekunnissa, kun suoritus toteutettiin edullisella yhden piirilevyn tietokoneella. Hieman tarkempia havainnointituloksia tuottaneet vertaillut menetelmät pystyivät käsittelemään vain muutamia kuvia sekunnissa. Kehitetty havainnointimenetelmä tarjoaa tehokkaan ratkaisun kameroissa reunalaskennalla toteutettaviin tunnistustehtäviin. Liikennekameroihin perustuvan tienkäyttäjien paikannuksen herkkyyttä tutkittiin, kvantifioiden erinäisten mittaustapahtuman virheiden vaikutuksia paikannustuloksiin. Tienkäyttäjien havainnoinnissa tapahtuvien lievien virheiden osoitettiin voivan muodostaa huomattavia virheitä paikannuksessa. Lisäksi kameran asentoon liittyvien virheiden havaittiin tuottavan paikannustuloksiin merkittäviä ongelmia. Tarkka kalibrointi on oleellista paikannussovelluksissa, ja kalibrointi täytyisi pystyä säilyttämään todenmukaisena. Automaattinen kalibrointimenetelmä kehitettiin liikennekameroille. Menetelmä perustuu verkottuneiden ajoneuvojen jakamaan satelliittipaikannustietoon. Aiempiin menetelmiin verrattuna kehitetty algoritmi on laajemmin sovellettavissa, sisältää konkreettisen takaisinkytkennän kalibrointituloksen arviointiin ja kalibroi kamerat globaaliin koordinaattijärjestelmään. Luodun menetelmän toimivuus todennettiin käytännön kokeilla. Tulevaisuuden tutkimustyön tulisi jatkaa menetelmien kehittämistä yhä toimintavarmemmiksi erilaisissa ympäristöissä. Liikennekamerat valvovat huomattavan erilaisia tieympäristöjä, joissa liikennetiheys, tienkäyttäjien jakauma ja kuvakulma voivat olla hyvin erilaisia. Liikenteen analysointiin suunniteltujen tietokonenäköalgoritmien tulisi toimia luotettavasti kaikissa esiintyvissä olosuhteissa, jotta niiden tuottamaa tietoa voitaisiin tehokkaasti hyödyntää älykkäissä liikennejärjestelmissä.

Julkaisun otsikon käännös	Tietokonenäön menetelmiä tienkäyttäjien tunnistukseen ja paikannukseen älykkäiden liikennejärjestelmien infrastruktuurissa
Alkuperäiskieli	Englanti
Pätevyys	Tohtorintutkinto
Myöntävä instituutio	Aalto-yliopisto
Valvoja/neuvonantaja	Tammi, Kari, Vastuuprofessori
Kustantaja	Aalto University
Painoksen ISBN	978-952-64-1326-6
Sähköinen ISBN	978-952-64-1327-3
Tila	Julkaistu - 2023
OKM-julkaisutyyppi	G5 Artikkeliväitöskirja

Tutkimusalat

tietokonenäkö
älykkäät liikennejärjestelmät

Pääsy asiakirjaan

http://urn.fi/URN:ISBN:978-952-64-1327-3

Siteeraa tätä

@phdthesis{ea712b594d7141e68513a9584892811e,

title = "Computer Vision for Road User Detection and Localisation in Intelligent Transportation System Infrastructure",

abstract = "Demand for intelligent transportation systems is on the rise. Intelligent features have the potential to solve numerous problems related to congestion, pollution, as well as safety of transportation. In order to implement different intelligent transport system applications, a great amount of real-time traffic information is needed for robust and reliable decision-making. This thesis explores solutions for utilising roadside cameras as real-time data acquisition tools for smart traffic applications. The research of this thesis was focused on solving problems related to practical applicability of computer vision for road user detection and localisation in roadside camera views. Experimental testing was carried out to validate the proposed methods, as well as produce results highlighting the capabilities of computer vision solutions for intelligent road user monitoring applications. A fast road user detection algorithm based on fusion of background subtraction and a convolutional neural network classifier was proposed. The detector was shown capable of running at over 30 frames per second on a low-cost single-board computer. The accuracy of the detector was only slightly lower than those of compared state-of-the-art object detectors that performed detection at single digit processing speeds. The developed detector is effective in edge computing applications that perform real-time road user detection. The sensitivity of roadside camera-based road user localisation was studied by quantifying errors caused by different inaccuracies in the measurement. The experiments highlighted that minor errors in road user detection can propagate in to large errors in the localisation. Furthermore, slight deviation in the camera extrinsic parameters can render the localisation results unusable. Careful calibration of a roadside camera is required for reliable localisation results. An automated calibration approach for roadside cameras was proposed, based on utilisation of global navigation satellite system positioning information of connected vehicles. Compared to previous automated methods, the proposed method is more generalisable, includes a feedback loop for verifying the calibration, and calibrates cameras to a global coordinate system. Experimental testing was carried out to validate the performance of the method. Future development should focus on further tuning the proposed methods with increased robustness for a wider variety of operating environments. Roadside cameras monitor highly different road environments with different types of traffic. Camera views can have immense variation in road user types, traffic density, as well as appearance and orientation of road users. Extensively robust algorithms are required to successfully deploy roadside computer vision solutions in intelligent transportation applications.",

keywords = "computer vision, intelligent transportation systems, tietokonen{\"a}k{\"o}, {\"a}lykk{\"a}{\"a}t liikennej{\"a}rjestelm{\"a}t, computer vision, intelligent transportation systems",

author = "Risto Ojala",

year = "2023",

language = "English",

isbn = "978-952-64-1326-6",

series = "Aalto University publication series DOCTORAL THESES",

publisher = "Aalto University",

number = "98",

address = "Finland",

school = "Aalto University",

}

TY - BOOK

T1 - Computer Vision for Road User Detection and Localisation in Intelligent Transportation System Infrastructure

AU - Ojala, Risto

PY - 2023

Y1 - 2023

N2 - Demand for intelligent transportation systems is on the rise. Intelligent features have the potential to solve numerous problems related to congestion, pollution, as well as safety of transportation. In order to implement different intelligent transport system applications, a great amount of real-time traffic information is needed for robust and reliable decision-making. This thesis explores solutions for utilising roadside cameras as real-time data acquisition tools for smart traffic applications. The research of this thesis was focused on solving problems related to practical applicability of computer vision for road user detection and localisation in roadside camera views. Experimental testing was carried out to validate the proposed methods, as well as produce results highlighting the capabilities of computer vision solutions for intelligent road user monitoring applications. A fast road user detection algorithm based on fusion of background subtraction and a convolutional neural network classifier was proposed. The detector was shown capable of running at over 30 frames per second on a low-cost single-board computer. The accuracy of the detector was only slightly lower than those of compared state-of-the-art object detectors that performed detection at single digit processing speeds. The developed detector is effective in edge computing applications that perform real-time road user detection. The sensitivity of roadside camera-based road user localisation was studied by quantifying errors caused by different inaccuracies in the measurement. The experiments highlighted that minor errors in road user detection can propagate in to large errors in the localisation. Furthermore, slight deviation in the camera extrinsic parameters can render the localisation results unusable. Careful calibration of a roadside camera is required for reliable localisation results. An automated calibration approach for roadside cameras was proposed, based on utilisation of global navigation satellite system positioning information of connected vehicles. Compared to previous automated methods, the proposed method is more generalisable, includes a feedback loop for verifying the calibration, and calibrates cameras to a global coordinate system. Experimental testing was carried out to validate the performance of the method. Future development should focus on further tuning the proposed methods with increased robustness for a wider variety of operating environments. Roadside cameras monitor highly different road environments with different types of traffic. Camera views can have immense variation in road user types, traffic density, as well as appearance and orientation of road users. Extensively robust algorithms are required to successfully deploy roadside computer vision solutions in intelligent transportation applications.

AB - Demand for intelligent transportation systems is on the rise. Intelligent features have the potential to solve numerous problems related to congestion, pollution, as well as safety of transportation. In order to implement different intelligent transport system applications, a great amount of real-time traffic information is needed for robust and reliable decision-making. This thesis explores solutions for utilising roadside cameras as real-time data acquisition tools for smart traffic applications. The research of this thesis was focused on solving problems related to practical applicability of computer vision for road user detection and localisation in roadside camera views. Experimental testing was carried out to validate the proposed methods, as well as produce results highlighting the capabilities of computer vision solutions for intelligent road user monitoring applications. A fast road user detection algorithm based on fusion of background subtraction and a convolutional neural network classifier was proposed. The detector was shown capable of running at over 30 frames per second on a low-cost single-board computer. The accuracy of the detector was only slightly lower than those of compared state-of-the-art object detectors that performed detection at single digit processing speeds. The developed detector is effective in edge computing applications that perform real-time road user detection. The sensitivity of roadside camera-based road user localisation was studied by quantifying errors caused by different inaccuracies in the measurement. The experiments highlighted that minor errors in road user detection can propagate in to large errors in the localisation. Furthermore, slight deviation in the camera extrinsic parameters can render the localisation results unusable. Careful calibration of a roadside camera is required for reliable localisation results. An automated calibration approach for roadside cameras was proposed, based on utilisation of global navigation satellite system positioning information of connected vehicles. Compared to previous automated methods, the proposed method is more generalisable, includes a feedback loop for verifying the calibration, and calibrates cameras to a global coordinate system. Experimental testing was carried out to validate the performance of the method. Future development should focus on further tuning the proposed methods with increased robustness for a wider variety of operating environments. Roadside cameras monitor highly different road environments with different types of traffic. Camera views can have immense variation in road user types, traffic density, as well as appearance and orientation of road users. Extensively robust algorithms are required to successfully deploy roadside computer vision solutions in intelligent transportation applications.

KW - computer vision

KW - intelligent transportation systems

KW - tietokonenäkö

KW - älykkäät liikennejärjestelmät

KW - computer vision

KW - intelligent transportation systems

M3 - Doctoral Thesis

SN - 978-952-64-1326-6

T3 - Aalto University publication series DOCTORAL THESES

PB - Aalto University

ER -

Computer Vision for Road User Detection and Localisation in Intelligent Transportation System Infrastructure

Abstrakti

Tutkimusalat

Pääsy asiakirjaan

Sormenjälki

Siteeraa tätä