Publication Date

2016

Document Type

Thesis

Committee Members

Paul Bender (Committee Member), Yong Pei (Advisor), Mateen Rizki (Committee Member)

Degree Name

Master of Science (MS)

Abstract

In this thesis, we have designed and developed a motion tracking and visualization system using the latest motion tracking sensory technologies. It is one of the enabling technologies for our novel visual-inertial odometer and human anatomy based 3D Locating, Mapping and Navigation system for endoscopy and drug delivery capsules used inside GI tract. In particular, we have: i) designed and completed a cloud-based sensory data collecting, processing and storage system to provide the reliable computing and storage platform; ii) explored different data processing methods to obtain improved-quality motion results from extremely noisy raw data, e.g., by using a low pass and high pass filter, and Kalman filters; iii) developed low-complexity algorithms to support real-time data analysis; and, iv) provided real-time 3 dimensional visualizations by a Unity 3D based visualizer. In this thesis, we have also showcased the use of application processors, which are widely used in smartphones and tablets, to develop a potentially low-cost sensor system and networks with enhanced computing, storage and networking capabilities. Specifically, we have explored Android/Android Wear, Google’s open source mobile OS’s, enabled smart devices, such as Sony Smartwatch 3, and their built-in sensory capabilities to build our sensory system. The completeness and maturity of such a widely used mobile platform ensure a fast prototype design and development process, as well as significantly better reliability and usability. The reliability of our sensory system is further improved through the use of a database approach, such as transactions, for the data flow from sensors to mobile platform and eventually to the cloud. Thus, our prototype design provides a working model to collect sensor data from Android Wear, and then transfer and store them into the cloud for further processing and disseminations.

Page Count

100

Department or Program

Department of Computer Science and Engineering

Year Degree Awarded

2016

Download


Share

COinS