High-Resolution Low-Cost Ultrasonic Tracking System for Human-Interface Systems

This work presents a low-speed 3-D position measurement system; designed for tracking simultaneously several target points of the human body. The target's positions are estimated based on acoustic time-difference-of-arrival (TDOA) estimations from different transmitters, which eliminates the effects of changes in the speed of sound. We propose a new signal-processing approach for this system, which is divided in two main parts. In the first part (mobile part on the human body), the signal processing is optimized for low-power consumption, low-sampling frequency, and only 154 samples, reducing the bandwidth required for the signal acquisition. In the second part (fixed part on personal computer), the signal processing estimates the TDOA using different techniques as Hilbert transform, zero-crossing detection, parabolic interpolation, cross-correlation and post-filter analysis. With a minimum number of data yet a high resolution is obtained using the proposed signal processing