Date of Award
Spring 1-1-2014
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Electrical, Computer & Energy Engineering
First Advisor
Nikolaus Correll
Second Advisor
Richard Han
Third Advisor
Dirk C. Grunwald
Abstract
Distributed Gesture recognition rather than centralized processing is of importance in distributed sensor arrays embedded within networked systems. We have implemented a gesture recognition algorithm on a smart wall system with distributed sensing using capacitive touch sensors. Each node records a sense event and shares it with other nodes in the network. Then each node builds up a chain vector array and compares it with a reference gestures to find out the alphabet drawn on the wall. We studied the effect of gesture speed and number of vectors on gesture recognition and found that as the number of vectors in a gesture increases, the recognition becomes more reliable. A slower gesture speed allows for better sensing of touch and hence also leads to better gesture recognition.
We also implemented a Hardware abstraction Architecture based on the TinyOS operating system to achieve the goal of making the program platform independent.
Recommended Citation
Mysore, Akshay Raghunandan, "Spatio-Temporal Gesture Recognition on a Distributed Sensing System" (2014). Electrical, Computer & Energy Engineering Graduate Theses & Dissertations. 90.
https://scholar.colorado.edu/ecen_gradetds/90