Date of Graduation

5-2018

Document Type

Thesis

Degree Name

Master of Science in Computer Science (MS)

Degree Level

Graduate

Department

Computer Science & Computer Engineering

Advisor/Mentor

Qinghua Li

Committee Member

Susan Gauch

Second Committee Member

Christophe Bobda

Keywords

Accelerometer, Home Security, iPhone, Machine Learning, Sensors, Spectrogram

Abstract

Several new smartphones are released every year. Many people upgrade to new phones, and their old phones are not put to any further use. In this paper, we explore the feasibility of using such retired smartphones and their on-board sensors to build a home security system. We observe that door-related events such as opening and closing have unique vibration signatures when compared to many types of environmental vibrational noise. These events can be captured by the accelerometer of a smartphone when the phone is mounted on a wall near a door. The rotation of a door can also be captured by the magnetometer of a smartphone when the phone is mounted on a door. We design machine learning and threshold-based methods to detect door opening events based on accelerometer and magnetometer data and build a prototype home security system that can detect door openings and notify the homeowner via email, SMS and phone calls upon break-in detection. To further augment our security system, we explore using the smartphone’s built-in microphone to detect door and window openings across multiple doors and windows simultaneously. Experiments in a residential home show that the accelerometer- based detection can detect door open events with an accuracy higher than 98%, and magnetometer-based detection has 100% accuracy. By using the magnetometer method to automate the training phase of a neural network, we find that sound-based detection of door openings has an accuracy of 90% across multiple doors.

Share

COinS