Video processing road safety system

Birkett, Russell Johnsen (2016) Video processing road safety system. [USQ Project]

[img]
Preview
Text
Birkett_R_LOW.pdf

Download (2455Kb) | Preview

Abstract

The purpose of this report was to investigate the possible applications of computer vision processing in roadway safety. With the huge expansion of computer vision software as well as affordable processing hardware opportunity to apply established methods to new areas has arisen.

A literature review was undertaken investigating and comparing algorithms in computer vision and their possible application for roadway safety. Through the literature review it was decided the most effective application for pedestrian tracking would be a combination of Gaussian Mixture Modeling for background subtraction and a Kalman filter for tracking. This in conjunction with a range of morphological filters was the final design decision. This design was cemented as a Simulink model and was run against a series of scenarios. The most important results were testing against a dynamic roadway showing extremely promising outcomes. The model was able to detect and track pedestrians over a range of designated areas. This in addition to its ability to differentiate between cars and pedestrians and the location of those pedestrians in relation to the road made the system a success. With simple counting and regions of interest changes the model can be applied to range of different environments and scenarios. However problems did emerge in the implementation of the Kalman filter causing issues in its application and track association. This report was dedicated to investigating and applying current methods and techniques in computer vision that could benefit pedestrian safety, in this regard it has successfully represented the possible uses and application of such a model.


Statistics for USQ ePrint 31375
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Additional Information: Bachelor of Engineering (Honours)
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Mechanical and Electrical Engineering
Supervisors: Low, Tobias
Date Deposited: 31 Jul 2017 02:39
Last Modified: 06 Dec 2017 01:10
Uncontrolled Keywords: roadway safety; computer vision; video processing; safety system
Fields of Research : 09 Engineering > 0913 Mechanical Engineering > 091302 Automation and Control Engineering
URI: http://eprints.usq.edu.au/id/eprint/31375

Actions (login required)

View Item Archive Repository Staff Only