Hardware Design Optimization for Human Motion Tracking Systems

ISBN-13: 9783639137255 / Angielski / Miękka / 2009 / 196 str.

B. Danette Allen

Opis

This research introduces a stochastic framework forevaluating and comparing the expected performance ofsensing systems for interactive computer graphics.Incorporating models of the sensor devices andexpected user motion dynamics, this framework enablescomplementary system- and measurement-level hardwareinformation optimization, independent of algorithmand motion paths. The approach for system-leveloptimization is to estimate the asymptotic positionand/or orientation uncertainty at many pointsthroughout a desired working volume or surface, andto visualize the results graphically. This globalperformance estimation can provide both a quantitative assessment of the expected performanceand intuition about how to improve the type andarrangement of sources and sensors, in the context ofthe desired working volume and expected scenedynamics. Using the same model components requiredfor these system-level optimization, the optimalsensor sampling time can be determined with respectto the expected scene dynamics for measurement-leveloptimization.

O autorze

Dr. B. Danette Allen is a senior researcher at NASA Langley
Research Center. She has extensive experience in the design and
development of atmospheric science instruments and is
investigating methods for modernizing the National Airspace
System. Dr. Allen received her Ph.D. in Computer Science from the
University of North Carolina at Chapel Hill.