COMING SOON! PQDT Open is getting a new home!

ProQuest Open Access Dissertations & Theses will remain freely available as part of a new and enhanced search experience at

Questions? Please refer to this FAQ.

Dissertation/Thesis Abstract

Acoustic and Image Vehicle Classification System
by Hoang, Trong Tuyen, M.S., California State University, Long Beach, 2017, 56; 10604463
Abstract (Summary)

The objective of this project is to detect and classify vehicles into two categories – 1) light vehicles, such as cars and other passenger vehicles and 2) heavy vehicles: trucks, busses, and other commercial vehicles, using video, sound, as well as joint video and sound classifications.

For the video portion, images from a video were analyzed. The physical size of a vehicle was extracted and stored. From this, the classification of different vehicles was made possible. In regards to the sound portion, the vehicles were recorded acoustically and their sound signature was analyzed. From these sound waves, it is desirable to extract not only the energy levels, wideband spectrogram, and converting it to the frequency domain, but to also represent it on a nonlinear mel scale of frequency. This has been frequently done for automatic speech recognition, but it is also feasible to extract the mel-frequency cepstral coefficients to use as features to aid the classification process of vehicles. Using this, one can train a machine learning algorithm - such as an artificial neural network - and have it label vehicles as one of the two classes.

Furthermore, the combination of both the video and the sound portion makes it possible to detect and classify vehicles more accurately than if using each system separately. This potentially allows for a more reliable system, especially when the image classification system by itself is not able to properly detect or classify vehicles, i.e. in heavy fog, or at night.

Indexing (document details)
Advisor: Khoo, I-Hung
Commitee: Chassiakos, Anastasios, Yeh, Hen-Geul
School: California State University, Long Beach
Department: Electrical Engineering
School Location: United States -- California
Source: MAI 56/06M(E), Masters Abstracts International
Subjects: Electrical engineering
Keywords: Classification, Detection, Machine learning, Neural network, Recognition, Vehicle
Publication Number: 10604463
ISBN: 978-0-355-23026-0
Copyright © 2021 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy