Please use this identifier to cite or link to this item: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4244
Title: Traffic prediction using vehicle inflow data captured by sensors
Authors: Kulatunge, M.H
Issue Date: 27-Jul-2021
Abstract: As the problem of urban traffic congestion spreads, there arises the need for the use of advanced technology to provide information about traffic congestion. Although many types of traffic sensors are currently in use, all have some drawbacks, and the deployment of such sensor systems has been difficult due to high costs. Sensors such as Inductive loop detections, Infrared sensors, Magnetic sensors are hard to deploy and inaccurate when comes to detecting traffic. Ultra sonic sensors and video cameras are used in this research due to the reasons that ultrasonic sensors are cost effective, easy to deploy and video cameras can be easily installed and supports scalability. Within the scope of this research, three major contributions are presented. First is an approach to detect traffic by capturing vehicle speed using ultra sonic sensors. The second contribution is using video processing to detect traffic of urban routes in Colombo by capturing vehicle speed and count. As the third contribution, a machine learning approach was used to detect traffic in urban routes of Colombo. Moreover, this research has used a statistical based approach to predict traffic using traffic data captured by the machine learning approach. Experiments to verify the suitability of ultra sonic sensors to capture vehicle speed was carried out in a prototype environment. Whereas the evaluation of the video processing to capture vehicle speed and count was carried out in a real world environment. For the machine learning approach a model was trained with over 4000 images and evaluated the accuracy of traffic detection with video feed taken from urban areas of Colombo.
URI: http://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4244
Appears in Collections:2018

Files in This Item:
File Description SizeFormat 
2015MCS041.pdf3.03 MBAdobe PDFView/Open


Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.