Handwriting Tracking using 60 GHz mmWave Radar

Abstract

Human-computer interaction is a vital component in today's world and there is a constant quest for automated and user-friendly techniques for interaction. Handwriting is one of the most general and natural ways of interaction for humans. While handwriting recognition is a well-studied problem, its counterpart handwriting tracking is still being investigated. Most of the existing handwriting tracking systems either require sensors attached to the hand or involve specially designed hardware. In this work, we propose a handwriting tracking system that reuses a commodity 60GHz Wi-Fi radio as a radar. The moving target is localized at each time instance and a trajectory is constructed by connecting those location estimates. While the digital beamforming technique and the pulsed radar are used to recover the spatial and range information, Doppler velocity is used to isolate the moving target from the static objects. Further, subsample peak interpolation and smoothing techniques enhance the overall performance of the proposed system. Extensive experiments demonstrate an average tracking error of 2.5% of the distance from the device and validate the robustness of the system to different environments and experimental conditions.