From Communication to Sensing : Recognizing and Counting Repetitive Motions with Wireless Backscattering

Abstract

Recently several ground-breaking RF-based motion recognition systems were proposed to detect and/or recognize macro/micro human movements. These systems often suffer from various interferences caused by multiple-users moving simultaneously, resulting in extremely low recognition accuracy. To tackle this challenge, we propose a novel system, called Motion-Fi, which marries battery free wireless backscattering and device-free sensing. Motion-Fi is an accurate, interference tolerable motion-recognition system, which counts repetitive motions without using scenario-dependent templates or profiles and enables multi-users performing certain motions simultaneously because of the relatively short transmission range of backscattered signals. Although the repetitive motions are fairly well detectable through the backscattering signals in theory, in reality they get blended into various other system noises during the motion. Moreover, irregular motion patterns among users will lead to expensive computation cost for motion recognition. We build a backscattering wireless platform to validate our design in various scenarios for over 6 months when different persons, distances and orientations are incorporated. In our experiments, the periodicity in motions could be recognized without any learning or training process, and the accuracy of counting such motions can be achieved within 5% count error. With little efforts in learning the patterns, our method could achieve 93.1% motion-recognition accuracy for a variety of motions. Moreover, by leveraging the periodicity of motions, the recognition accuracy could be further improved to nearly 100% with only 3 repetitions. Our experiments also show that the motions of multiple persons separating by around 2 meters cause little accuracy reduction in the counting process.