Dissertation/Thesis Abstract

In-Node Vehicle Classification and Identification
by Bavkar, Indraneel N., M.S., California State University, Long Beach, 2017, 50; 10639676
Abstract (Summary)

The congestion on the freeways as well as on the city roads is increasing day by day due to a continuous rise in the traffic. Better traffic systems using advanced intelligent components are urgently needed to manage this continuous growth of traffic. Currently, traffic departments in the United States utilize inductive loop technologies to provide a solution to this problem by detecting the cars to control the traffic lights. These inductive loops are placed under the road during the construction of the road, or they are placed in the road at a later point by sawing through the surface. As stated earlier, these inductive loops are primarily used for the detection of vehicles and they consume a large amount of power. What if vehicles could be classified as well as identified instead of just being detected? This would make the traffic control system much more advanced and intelligent. To achieve this, an in-node microprocessor-based vehicle classification approach is proposed in this thesis. It can analyze the vehicles’ magnetic data and determine the types of vehicles passing over a 3-axis magnetometer sensor. This is done using the J48 classification algorithm, which is implemented in Waikato Environment for Knowledge Analysis (WEKA), a machine learning software suite, and the Naïve Bayes model implemented in MATLAB (MATrix LABoratory) as well. The J48 is a decision tree machine learning algorithm derived from Quinlan's C4.5 algorithm which is based on the ID3 (Iterative Dichotomiser 3) algorithm. Features are extracted from the data collected from vehicles passing over the sensor and a decision tree model is generated based on those features. The decision tree iii model that is generated is then implemented using decision commands such as if-else in any language and on any microprocessor platform. Also, to make the sensor node reusable at different locations and in different environments, an adaptive baseline is set up to eliminate the background magnetic field from the raw data. Binary Naïve Bayes model is also used in addition to the J48 algorithm as only two major classes of vehicles are being analyzed which are Sedan and Hatchback.

Indexing (document details)
Advisor: Mozumdar, Mohammad
Commitee: Ahmed, Aftab, Kwon, Seok-Chul (Sean)
School: California State University, Long Beach
Department: Electrical Engineering
School Location: United States -- California
Source: MAI 57/01M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Electrical engineering, Transportation, Artificial intelligence
Keywords: Decision tree, Machine learning, Magnetometer, Neural network
Publication Number: 10639676
ISBN: 978-0-355-52951-7
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest