Computer Vision Algorithms for Intelligent Transportation Systems Applications

• In recent years, Intelligent Transportation Systems (ITS) have emerged asan efficient way of enhancing traffic flow, safety and management. Thesegoals are realized by combining various technologies and analyzing the acquireddata from vehicles and roadways. Among all ITS technologies, computervision solutions have the advantages of high flexibility, easy maintenanceand high price-performance ratio that make them very popular fortransportation surveillance systems. However, computer vision solutionsare demanding and challenging due to computational complexity, reliability,efficiency and accuracy among other aspects. In this thesis, three transportation surveillance systems based on computervision are presented. These systems are able to interpret the imagedata and extract the information about the presence, speed and class ofvehicles, respectively. The image data in these proposed systems are acquiredusing Unmanned Aerial Vehicle (UAV) as a non-stationary sourceand roadside camera as a stationary source. The goal of these works is toenhance the general performance of accuracy and robustness of the systemswith variant illumination and traffic conditions. This is a compilation thesis in systems engineering consisting of threeparts. The red thread through each part is a transportation surveillancesystem. The first part presents a change detection system using aerial imagesof a cargo port. The extracted information shows how the space isutilized at various times aiming for further management and developmentof the port. The proposed solution can be used at different viewpoints andillumination levels e.g. at sunset. The method is able to transform the imagestaken from different viewpoints and match them together. Thereafter,it detects discrepancies between the images using a proposed adaptive localthreshold. In the second part, a video-based vehicle's speed estimationsystem is presented. The measured speeds are essential information for lawenforcement and they also provide an estimation of traffic flow at certainpoints on the road. The system employs several intrusion lines to extractthe movement pattern of each vehicle (non-equidistant sampling) as an inputfeature to the proposed analytical model. In addition, other parameters such as camera sampling rate and distances between intrusion lines are alsotaken into account to address the uncertainty in the measurements and toobtain the probability density function of the vehicle's speed. In the thirdpart, a vehicle classification system is provided to categorize vehicles into\private car", \light trailer", \lorry or bus" and \heavy trailer". This informationcan be used by authorities for surveillance and development ofthe roads. The proposed system consists of multiple fuzzy c-means clusterings using input features of length, width and speed of each vehicle. Thesystem has been constructed by using prior knowledge of traffic regulationsregarding each class of vehicle in order to enhance the classification performance.

Ämnesord

TEKNIK OCH TEKNOLOGIER  -- Elektroteknik och elektronik -- Signalbehandling (hsv//swe)

ENGINEERING AND TECHNOLOGY  -- Electrical Engineering, Electronic Engineering, Information Engineering -- Signal Processing (hsv//eng)

NATURVETENSKAP  -- Data- och informationsvetenskap -- Datorseende och robotik (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences -- Computer Vision and Robotics (hsv//eng)

NATURVETENSKAP  -- Data- och informationsvetenskap -- Annan data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences -- Other Computer and Information Science (hsv//eng)

Nyckelord

computer vision

intelligent transportation systems (ITS)

speed measurement

vehicle classification

Publikations- och innehållstyp

vet (ämneskategori)

lic (ämneskategori)