Geo-temporal Online Analysis of Traffic Rule Violations

University essay from Göteborgs universitet/Institutionen för data- och informationsteknik

Abstract: Due to inattention and not complying with traffic regulations, human error accounts forroughly 94% of all traffic accidents. To counter this, the need to develop systems that canidentify traffic rule violations and calculate the risk of collisions. The information reportedcan then be used to implement preventive measures. Modern vehicles are equipped withsensors and cameras thus making this possible, but it comes with the complication of notviolating the privacy of individuals when gathering information.This project presents a prototype system comprised of three subsystems with the intentionof reducing traffic accidents. The first two revolve around the detection of traffic violationswith the use of real-time object detection and intention aware risk estimation. Thepurpose of the third subsystem is to detach personal information from the data gatheredby the previously mentioned subsystems. This makes it possible to use the data topinpoint problematic areas in a traffic environment.Evaluation of the system was performed in both a simulation environment and withanalysis of video feeds from a lab environment. The results of the evaluation showthat the prototype system developed in the project is sufficiently accurate to be furtherdeveloped and implemented for use in real vehicles.

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)