Developing a Vehicle Categorizer Solution

OData support
Supervisor:
Simon Gábor
Department of Automation and Applied Informatics

The scope of this dissertation is to propose a solution for analyzing vehicles involved in traffic and determine of which category they belong to (passenger or lorry vehicle). This information allows the detection of such vehicle category dependent violations, as speeding, forbidden overtaking, forbidden entry/exit and irregular parking.

There are a number of categorization algorithms, successfully utilized in current industrial solutions, but most of them are not designed for resource-critical environments, and to develop an implementation that runs in nearly real-time on an Intel Atom is still an area of active research.

The thesis discusses

• what is expected from the selected algorithm,

• provides an insight into methods and technologies used today at the field of image processing,

• describes and analyzes the possible solutions,

• outlines the resource constraints for running the algorithm,

• justifies the need for using supplementary sensor data,

• details the steps of design and implementation,

• summarizes the results, and suggests a direction for further development based on the acquired experience.

Downloads

Please sign in to download the files of this thesis.